Journal articles on the topic 'Lesion detection'
To see the other types of publications on this topic, follow the link: Lesion detection.
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 journal articles for your research on the topic 'Lesion detection.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
Zrzavy, Tobias, Alice Wielandner, Lukas Haider, Sophie Bartsch, Fritz Leutmezer, Thomas Berger, Karl Heinz Nenning, Alexander Rauscher, Paulus Rommer, and Gregor Kasprian. "FLAIR2 post-processing: improving MS lesion detection in standard MS imaging protocols." Journal of Neurology 269, no. 1 (October 8, 2021): 461–67. http://dx.doi.org/10.1007/s00415-021-10833-x.
Full textAbstract:
Abstract Background Technical improvements in magnetic resonance imaging (MRI) acquisition, such as higher field strength and optimized sequences, lead to better multiple sclerosis (MS) lesion detection and characterization. Multiplication of 3D-FLAIR with 3D-T2 sequences (FLAIR2) results in isovoxel images with increased contrast-to-noise ratio, increased white–gray-matter contrast, and improved MS lesion visualization without increasing MRI acquisition time. The current study aims to assess the potential of 3D-FLAIR2 in detecting cortical/leucocortical (LC), juxtacortical (JC), and white matter (WM) lesions. Objective To compare lesion detection of 3D-FLAIR2 with state-of-the-art 3D-T2-FLAIR and 3D-T2-weighted images. Methods We retrospectively analyzed MRI scans of thirteen MS patients, showing previously noted high cortical lesion load. Scans were acquired using a 3 T MRI scanner. WM, JC, and LC lesions were manually labeled and manually counted after randomization of 3D-T2, 3D-FLAIR, and 3D-FLAIR2 scans using the ITK-SNAP tool. Results LC lesion visibility was significantly improved by 3D-FLAIR2 in comparison to 3D-FLAIR (4 vs 1; p = 0.018) and 3D-T2 (4 vs 1; p = 0.007). Comparing LC lesion detection in 3D-FLAIR2 vs. 3D-FLAIR, 3D-FLAIR2 detected on average 3.2 more cortical lesions (95% CI − 9.1 to 2.8). Comparing against 3D-T2, 3D-FLAIR2 detected on average 3.7 more LC lesions (95% CI 3.3–10.7). Conclusions 3D-FLAIR2 is an easily applicable time-sparing MR post-processing method to improve cortical lesion detection. Larger sampled studies are warranted to validate the sensitivity and specificity of 3D-FLAIR2.
2
Correia, J., A. Ponte, L. Proença, A. Rodrigues, R. Pinho, S. Leite, C. Fernandes, et al. "P206 Comparison of dye-spraying chromoendoscopy and virtual chromoendoscopy for colonic dysplasia detection in longstanding Inflammatory Bowel Disease." Journal of Crohn's and Colitis 16, Supplement_1 (January 1, 2022): i265. http://dx.doi.org/10.1093/ecco-jcc/jjab232.333.
Full textAbstract:
Abstract Background Dye-spraying chromoendoscopy (DCE) is the technique of choice for colonic surveillance in patients with long-standing extensive Inflammatory Bowel Disease (IBD). Conversely, the use of virtual chromoendoscopy (VCE) is still controversial. This study aimed to compare lesion and dysplasia detection and accuracy of margins delineation between VCE and DCE. Methods Eleven gastroenterologists were given a survey with 20 pairs of pictures of IBD surveillance colonoscopies (10 with nondysplastic lesions, 5 with dysplastic lesions and 5 with no lesions). Each pair contained the same image captured during colonoscopy using VCE and DCE. The 40 pictures were randomly ordered to avoid any classification bias. For each picture, the gastroenterologist assessed the presence/absence of lesion and, when a lesion was identified, assessed the presence/absence of dysplasia and delineated its margins. To compare lesion and dysplasia detections between techniques, the sensitivity, specificity and inter-observer agreement (using fleiss’ kappa (K) test) were calculated. The chi-square test was used to assess the accuracy of margins delineation. Results When assessing lesion detection using VCE, sensitivity (S) and specificity (E) were 0.93 and 0.49 and in, DCE, 0.97 and 0.38, respectively. When assessing dysplasia detection using VCE, S and E were 0.74 and 0.60 and, in DCE, 0.67 and 0.62, respectively. Interobserver agreement analysis revealed that VCE and DCE had a moderate agreement in lesion detection - 0.57 and 0.58, respectively; however, for dysplasia detection, VCE had a fair agreement (k=0.30) and DCE a slight agreement (k=0.11). The rate of accurately defined margins was similar for both techniques (p=0.22). Conclusion Similar lesion and dysplasia detection and margins delineation were achieved with both techniques. However, concerning dysplasia detection, interobserver agreement was slightly better using VCE. Therefore, VCE may constitute a valid alternative to DCE for dysplasia screening in IBD. Further studies are needed to validate these findings.
3
Wattjes, Mike P., Martijn T. Wijburg, Anke Vennegoor, Birgit I. Witte, Stefan D. Roosendaal, Esther Sanchez, Yaou Liu, et al. "Diagnostic performance of brain MRI in pharmacovigilance of natalizumab-treated MS patients." Multiple Sclerosis Journal 22, no. 9 (July 20, 2016): 1174–83. http://dx.doi.org/10.1177/1352458515615225.
Full textAbstract:
Background: In natalizumab-treated multiple sclerosis (MS) patients, magnetic resonance imaging (MRI) is considered as a sensitive tool in detecting both MS disease activity and progressive multifocal leukoencephalopathy (PML). Objective: To investigate the performance of neuroradiologists using brain MRI in detecting new MS lesions and asymptomatic PML lesions and in differentiating between MS and PML lesions in natalizumab-treated MS patients. The secondary aim was to investigate interrater variability. Methods: In this retrospective diagnostic study, four blinded neuroradiologists assessed reference and follow-up brain MRI scans of 48 natalizumab-treated MS patients with new asymptomatic PML lesions ( n = 21) or new MS lesions ( n = 20) or no new lesions ( n = 7). Sensitivity and specificity for detection of new lesions in general (MS and PML lesions), MS and PML lesion differentiation, and PML detection were determined. Interrater agreement was calculated. Results: Overall sensitivity and specificity for the detection of new lesions, regardless of the nature of the lesions, were 77.4% and 89.3%, respectively; for PML-MS lesion differentiation, 74.2% and 84.7%, respectively; and for asymptomatic PML lesion detection, 59.5% and 91.7%, respectively. Interrater agreement for the tested categories was fair to moderate. Conclusion: The diagnostic performance of trained neuroradiologists using brain MRI in pharmacovigilance of natalizumab-treated MS patients is moderately good. Interrater agreement among trained readers is fair to moderate.
4
Metser, Ur, Orit Golan, Charles D. Levine, and Einat Even-Sapir. "Tumor Lesion Detection." Journal of Computer Assisted Tomography 29, no. 4 (July 2005): 554–59. http://dx.doi.org/10.1097/01.rct.0000164671.96143.c2.
Full text
5
Bhardwaj, Charu, Shruti Jain, and Meenakshi Sood. "Automated Diagnostic Hybrid Lesion Detection System for Diabetic Retinopathy Abnormalities." International Journal of Sensors, Wireless Communications and Control 10, no. 4 (December 18, 2020): 494–507. http://dx.doi.org/10.2174/2210327909666191126092411.
Full textAbstract:
Background: Early diagnosis, monitoring disease progression, and timely treatment of Diabetic Retinopathy (DR) abnormalities can efficiently prevent visual loss. A prediction system for the early intervention and prevention of eye diseases is important. The contrast of raw fundus image is also a hindrance in effective manual lesion detection technique. Methods: In this research paper, an automated lesion detection diagnostic scheme has been proposed for early detection of retinal abnormalities of red and yellow pathological lesions. The algorithm of the proposed Hybrid Lesion Detection (HLD) includes retinal image pre-processing, blood vessel extraction, optical disc localization and detection stages for detecting the presence of diabetic retinopathy lesions. Automated diagnostic systems assist the ophthalmologists practice manual lesion detection techniques which are tedious and time-consuming. Detailed statistical analysis is performed on the extracted shape, intensity and GLCM features and the optimal features are selected to classify DR abnormalities. Exhaustive statistical investigation of the proposed approach using visual and empirical analysis resulted in 31 significant features. Results: The results show that the HLD approach achieved good classification results in terms of three statistical indices: accuracy, 98.9%; sensitivity, 97.8%; and specificity, 100% with significantly less complexity. Conclusion: The proposed technique with optimal features demonstrates improvement in accuracy as compared to state of the art techniques using the same database.
6
Kuklyte, Jogile, Jenny Fitzgerald, Sophie Nelissen, Haolin Wei, Aoife Whelan, Adam Power, Ajaz Ahmad, et al. "Evaluation of the Use of Single- and Multi-Magnification Convolutional Neural Networks for the Determination and Quantitation of Lesions in Nonclinical Pathology Studies." Toxicologic Pathology 49, no. 4 (February 23, 2021): 815–42. http://dx.doi.org/10.1177/0192623320986423.
Full textAbstract:
Digital pathology platforms with integrated artificial intelligence have the potential to increase the efficiency of the nonclinical pathologist’s workflow through screening and prioritizing slides with lesions and highlighting areas with specific lesions for review. Herein, we describe the comparison of various single- and multi-magnification convolutional neural network (CNN) architectures to accelerate the detection of lesions in tissues. Different models were evaluated for defining performance characteristics and efficiency in accurately identifying lesions in 5 key rat organs (liver, kidney, heart, lung, and brain). Cohorts for liver and kidney were collected from TG-GATEs open-source repository, and heart, lung, and brain from internally selected R&D studies. Annotations were performed, and models were trained on each of the available lesion classes in the available organs. Various class-consolidation approaches were evaluated from generalized lesion detection to individual lesion detections. The relationship between the amount of annotated lesions and the precision/accuracy of model performance is elucidated. The utility of multi-magnification CNN implementations in specific tissue subtypes is also demonstrated. The use of these CNN-based models offers users the ability to apply generalized lesion detection to whole-slide images, with the potential to generate novel quantitative data that would not be possible with conventional image analysis techniques.
7
Matsui, Hiroaki, Shunsuke Kamba, Hideka Horiuchi, Sho Takahashi, Masako Nishikawa, Akihiro Fukuda, Aya Tonouchi, et al. "Detection Accuracy and Latency of Colorectal Lesions with Computer-Aided Detection System Based on Low-Bias Evaluation." Diagnostics 11, no. 10 (October 17, 2021): 1922. http://dx.doi.org/10.3390/diagnostics11101922.
Full textAbstract:
We developed a computer-aided detection (CADe) system to detect and localize colorectal lesions by modifying You-Only-Look-Once version 3 (YOLO v3) and evaluated its performance in two different settings. The test dataset was obtained from 20 randomly selected patients who underwent endoscopic resection for 69 colorectal lesions at the Jikei University Hospital between June 2017 and February 2018. First, we evaluated the diagnostic performances using still images randomly and automatically extracted from video recordings of the entire endoscopic procedure at intervals of 5 s, without eliminating poor quality images. Second, the latency of lesion detection by the CADe system from the initial appearance of lesions was investigated by reviewing the videos. A total of 6531 images, including 662 images with a lesion, were studied in the image-based analysis. The AUC, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 0.983, 94.6%, 95.2%, 68.8%, 99.4%, and 95.1%, respectively. The median time for detecting colorectal lesions measured in the lesion-based analysis was 0.67 s. In conclusion, we proved that the originally developed CADe system based on YOLO v3 could accurately and instantaneously detect colorectal lesions using the test dataset obtained from videos, mitigating operator selection biases.
8
Rezzo, R., G. Scopinaro, M. Gambaro, P. Michetti, and G. Anfossi. "Radioguided Occult Colonic Lesion Identification (Rocli) during Open and Laparoscopic Surgery." Tumori Journal 88, no. 3 (May 2002): S19—S22. http://dx.doi.org/10.1177/030089160208800328.
Full textAbstract:
Aims and Background Intraoperative localization, during open and laparoscopic surgery, of small, nonpalpable colonic lesions located at peculiar sites or with concurrent inflammatory bowel alterations (diverticulosis, perivisceritis) is often difficult. The aim of our work was to assess the validity of radioguided identification after preoperative labeling. Methods and Study Design Patients who were candidates for colon surgery for occult lesions that, because of their size and location, were assumed to be difficult to detect, underwent colonoscopy 1 to 2.5 hours before surgery. A small dose of labeled albumin macroaggregates was injected with a sclerotherapy needle into the subserosa underneath the lesion. Immediately following injection the lesion was identified with a transcutaneously placed gamma detecting probe. Intraoperative tracer detection was performed either during open surgery or by means of a laparoscopic probe (detection time 3-5 mins). The position of the lesion was marked with a suture or with a clip. Surgery was performed according to the type of lesion to be treated. Results In our initial clinical experience 15 colon lesions were preoperatively marked in 14 patients and were subsequently detected during surgery (four under laparoscopy) with a gamma detecting probe. This technique allows highly accurate, fast, and inexpensive surgical localization of lesions without irradiation and without complications. Conclusion Our experience shows that preoperative endoscopic marking of nonpalpable colon lesions with 99mTc-labeled albumin macroaggregates followed by intraoperative detection with a gamma probe is a useful clinical method that is highly accurate and without complications.
9
Sen Saxena, Vivek, Prashant Johri, and Avneesh Kumar. "AI-Enabled Support System for Melanoma Detection and Classification." International Journal of Reliable and Quality E-Healthcare 10, no. 4 (October 2021): 58–75. http://dx.doi.org/10.4018/ijrqeh.2021100104.
Full textAbstract:
Skin lesion melanoma is the deadliest type of cancer. Artificial intelligence provides the power to classify skin lesions as melanoma and non-melanoma. The proposed system for melanoma detection and classification involves four steps: pre-processing, resizing all the images, removing noise and hair from dermoscopic images; image segmentation, identifying the lesion area; feature extraction, extracting features from segmented lesion and classification; and categorizing lesion as malignant (melanoma) and benign (non-melanoma). Modified GrabCut algorithm is employed to generate skin lesion. Segmented lesions are classified using machine learning algorithms such as SVM, k-NN, ANN, and logistic regression and evaluated on performance metrics like accuracy, sensitivity, and specificity. Results are compared with existing systems and achieved higher similarity index and accuracy.
10
O'Brien, W. J., L. Vazquez, and W. M. Johnston. "The Detection of Incipient Caries with Tracer Dyes." Journal of Dental Research 68, no. 2 (February 1989): 157–58. http://dx.doi.org/10.1177/00220345890680021101.
Full textAbstract:
The purpose of this study was to determine the increase in color contrast produced by the use of a tracer dye in detection of incipient caries lesions with transillumination. Twenty four caries-free first premolars were immersed in an acid gelatin for production of artificial incipient caries lesions. After the lesions had developed, these teeth were photographed by transillumination. Two photographs were taken of each tooth. The first photograph showed the lesion without dye. A blue tracer dye was then added and absorbed by the lesion, and a second photograph was taken. The data on the color difference were obtained by use of a reflectance colorimeter and showed a four-fold increase between the lesion and surrounding area with the dye. A two-way analysis of variance was used for the statistical interpretation. The color difference between the lesion without the dye and then with the dye was significant. The use of the blue tracer dye, therefore, significantly increased the contrast in the images of the artificial incipient lesions.
11
Urzúa, I., R. Cabello, P. Marín, B. Ruiz, D. Jazanovich, C. Mautz, M. Lira, et al. "Detection of Approximal Caries Lesions in Adults: A Cross-sectional Study." Operative Dentistry 44, no. 6 (November 1, 2019): 589–94. http://dx.doi.org/10.2341/17-314-c.
Full textAbstract:
SUMMARY Detection and management of posterior approximal caries lesions are still problematic. Inspection of approximal surfaces is challenging, and bitewing radiographs are used when direct vision is not possible. Unfortunately, there is no definite radiographic appearance to identify lesion cavitation with absolute certainty. Many lesions detected radiographically within the outer half of dentin are not cavitated, often resulting in unnecessary restorative treatment. Our study compared radiographic depth of approximal caries lesions with presence of cavitation in adults using visual inspection following temporary tooth separation (TTS). We conducted this observational descriptive cross-sectional study at two dental schools in two cities in Chile. Clinicians were unaware of radiographic depths of lesions and examined 147 participants (57.3% female and 42.7% male) following TTS. Using the common classification system that consists of E0 (no lesion), E1 (lesion within the outer half of enamel), E2 (lesion within the inner half of enamel), D1 (lesion within the outer third of dentin), D2 (lesion within the middle third of dentin), and D3 (lesion within the inner third of dentin), a trained dentist evaluated all the processed films. Cavitation was detected in only three sites (0.22%) within the E0 category, seven sites (3.41%) in E1, five sites (14.8%) in E2, four sites (14.8%) in D1, six sites (50%) in D2, and eight sites (61.5%) in D3. Considering that restorative treatment should be indicated strictly for cavitated lesions, our findings support indication for restorative treatment for D3 lesions and the rationale for TTS for D1-D2 caries lesions to allow direct visual inspection to determine whether there is surface cavitation.
12
Ruohonen, Mika, Katri Palo, and Jarmo Alander. "Spectroscopic Detection of Caries Lesions." Journal of Medical Engineering 2013 (January 8, 2013): 1–9. http://dx.doi.org/10.1155/2013/161090.
Full textAbstract:
Background. A caries lesion causes changes in the optical properties of the affected tissue. Currently a caries lesion can be detected only at a relatively late stage of development. Caries diagnosis also suffers from high interobserver variance. Methods. This is a pilot study to test the suitability of an optical diffuse reflectance spectroscopy for caries diagnosis. Reflectance visible/near-infrared spectroscopy (VIS/NIRS) was used to measure caries lesions and healthy enamel on extracted human teeth. The results were analysed with a computational algorithm in order to find a rule-based classification method to detect caries lesions. Results. The classification indicated that the measured points of enamel could be assigned to one of three classes: healthy enamel, a caries lesion, and stained healthy enamel. The features that enabled this were consistent with theory. Conclusions. It seems that spectroscopic measurements can help to reduce false positives at in vitro setting. However, further research is required to evaluate the strength of the evidence for the method’s performance.
13
Kalinovsky, A., V. Liauchuk, and A. Tarasau. "LESION DETECTION IN CT IMAGES USING DEEP LEARNING SEMANTIC SEGMENTATION TECHNIQUE." ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLII-2/W4 (May 10, 2017): 13–17. http://dx.doi.org/10.5194/isprs-archives-xlii-2-w4-13-2017.
Full textAbstract:
In this paper, the problem of automatic detection of tuberculosis lesion on 3D lung CT images is considered as a benchmark for testing out algorithms based on a modern concept of Deep Learning. For training and testing of the algorithms a domestic dataset of 338 3D CT scans of tuberculosis patients with manually labelled lesions was used. The algorithms which are based on using Deep Convolutional Networks were implemented and applied in three different ways including slice-wise lesion detection in 2D images using semantic segmentation, slice-wise lesion detection in 2D images using sliding window technique as well as straightforward detection of lesions via semantic segmentation in whole 3D CT scans. The algorithms demonstrate superior performance compared to algorithms based on conventional image analysis methods.
14
Jiang, Chenhan, Shaoju Wang, Xiaodan Liang, Hang Xu, and Nong Xiao. "ElixirNet: Relation-Aware Network Architecture Adaptation for Medical Lesion Detection." Proceedings of the AAAI Conference on Artificial Intelligence 34, no. 07 (April 3, 2020): 11093–100. http://dx.doi.org/10.1609/aaai.v34i07.6765.
Full textAbstract:
Most advances in medical lesion detection network are limited to subtle modification on the conventional detection network designed for natural images. However, there exists a vast domain gap between medical images and natural images where the medical image detection often suffers from several domain-specific challenges, such as high lesion/background similarity, dominant tiny lesions, and severe class imbalance. Is a hand-crafted detection network tailored for natural image undoubtedly good enough over a discrepant medical lesion domain? Is there more powerful operations, filters, and sub-networks that better fit the medical lesion detection problem to be discovered? In this paper, we introduce a novel ElixirNet that includes three components: 1) TruncatedRPN balances positive and negative data for false positive reduction; 2) Auto-lesion Block is automatically customized for medical images to incorporates relation-aware operations among region proposals, and leads to more suitable and efficient classification and localization. 3) Relation transfer module incorporates the semantic relationship and transfers the relevant contextual information with an interpretable graph, thus alleviates the problem of lack of annotations for all types of lesions. Experiments on DeepLesion and Kits19 prove the effectiveness of ElixirNet, achieving improvement of both sensitivity and precision over FPN with fewer parameters.
15
Wang, L., S. Wang, S. Huang, and C. Liu. "DETECTING PROTRUSION LESION IN DIGESTIVE TRACT USING A SINGLE-STAGE DETECTION METHOD." ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLII-2/W12 (May 9, 2019): 231–35. http://dx.doi.org/10.5194/isprs-archives-xlii-2-w12-231-2019.
Full textAbstract:
<p><strong>Abstract.</strong> The classification networks have already existed for a long time and achieve great success. However, in biomedical image processing, classifying normal and abnormal ones only is not enough clinically, the desired output should include localization, i.e., where the lesion is. In this paper, we present a method for detecting protrusion lesion in digestive tract. We use a deep learning-based model to build a computer-aided diagnosis system to help doctors examine the intestinal diseases. Learn from existing detection method, one-stage and two-stage detection algorithm, a new network suitable for protrusion lesion detection is proposed. We inherit the method of anchor generation in SSD, a fast single-stage object detector outperform R-CNN series in terms of speed. Multi-scale feature layers are assigned to generate different sizes of default anchor boxes. Different from the previous work, our method doesnt require additional preprocessing because the network can learn features autonomously. For the 256*256 input, our method achieves 73% AP, perform a novel way to detect protrusion lesions.</p>
16
Madiwa, Shweta M., and Vishwanath Burkpalli. "Sine Cosine Based Harris Hawks Optimizer: A Hybrid Optimization Algorithm for Skin Cancer Detection Using Deep Stack Auto Encoder." Revue d'Intelligence Artificielle 36, no. 5 (December 23, 2022): 697–708. http://dx.doi.org/10.18280/ria.360506.
Full textAbstract:
Skin cancer is becoming major problems due to its tremendous growth. Skin cancer is a malignant skin lesion, which may cause damage to human. Hence, prior detection and precise medical diagnosis of the skin lesion is essential. In medical practice, detection of malignant lesions needs pathological examination and biopsy, which is expensive. The existing techniques need a brief physical inspection, which is imprecise and time-consuming. This paper presents a computer-assisted skin cancer detection strategy for detecting the skin lesion in skin images using deep stacked auto encoder. Sine Cosine-based Harris Hawks Optimizer (SCHHO) trains deep stacked auto encoders. The proposed SCHHO algorithm is designed by combining Sine Cosine Algorithm (SCA) and Harris Hawks Optimizer (HHO). The identification of skin lesion is performed on each segment, which is obtained by sparse-Fuzzy-c-means (FCM) algorithm. Statistical features, texture features and entropy are employed for selecting the most significant feature. Mean, standard deviation, variance, kurtosis, entropy, and Linear Discriminant Analysis (LDP) featured are extracted. SCHHO-Deep stacked auto-encoder outperformed other approaches with 91.66% accuracy, 91.60% sensitivity, and 91.72% specificity.
17
Xu, Lina, Giles Tetteh, Jana Lipkova, Yu Zhao, Hongwei Li, Patrick Christ, Marie Piraud, Andreas Buck, Kuangyu Shi, and Bjoern H. Menze. "Automated Whole-Body Bone Lesion Detection for Multiple Myeloma on 68Ga-Pentixafor PET/CT Imaging Using Deep Learning Methods." Contrast Media & Molecular Imaging 2018 (2018): 1–11. http://dx.doi.org/10.1155/2018/2391925.
Full textAbstract:
The identification of bone lesions is crucial in the diagnostic assessment of multiple myeloma (MM). 68Ga-Pentixafor PET/CT can capture the abnormal molecular expression of CXCR-4 in addition to anatomical changes. However, whole-body detection of dozens of lesions on hybrid imaging is tedious and error prone. It is even more difficult to identify lesions with a large heterogeneity. This study employed deep learning methods to automatically combine characteristics of PET and CT for whole-body MM bone lesion detection in a 3D manner. Two convolutional neural networks (CNNs), V-Net and W-Net, were adopted to segment and detect the lesions. The feasibility of deep learning for lesion detection on 68Ga-Pentixafor PET/CT was first verified on digital phantoms generated using realistic PET simulation methods. Then the proposed methods were evaluated on real 68Ga-Pentixafor PET/CT scans of MM patients. The preliminary results showed that deep learning method can leverage multimodal information for spatial feature representation, and W-Net obtained the best result for segmentation and lesion detection. It also outperformed traditional machine learning methods such as random forest classifier (RF), k-Nearest Neighbors (k-NN), and support vector machine (SVM). The proof-of-concept study encourages further development of deep learning approach for MM lesion detection in population study.
18
Moraal, Bastiaan, Stefan D. Roosendaal, Petra J. W. Pouwels, Hugo Vrenken, Ronald A. van Schijndel, Dominik S. Meier, Charles R. G. Guttmann, Jeroen J. G. Geurts, and Frederik Barkhof. "Multi-Contrast, Isotropic, Single-Slab 3D MR Imaging in Multiple Sclerosis." Neuroradiology Journal 22, no. 1_suppl (September 2009): 33–42. http://dx.doi.org/10.1177/19714009090220s108.
Full textAbstract:
To describe signal and contrast properties of an isotropic, single-slab 3D dataset [double inversion-recovery (DIR), fluid-attenuated inversion recovery (FLAIR), T2, and T1-weighted magnetization prepared rapid acquisition gradient-echo (MPRAGE)] and to evaluate its performance in detecting multiple sclerosis (MS) brain lesions compared to 2D T2-weighted spin-echo (T2SE). All single-slab 3D sequences and 2D-T2SE were acquired in 16 MS patients and 9 age-matched healthy controls. Lesions were scored independently by two raters and characterized anatomically. Two-tailed Bonferroni-corrected Student's t-tests were used to detect differences in lesion detection between the various sequences per anatomical area after log-transformation. In general, signal and contrast properties of the 3D sequences enabled improved detection of MS brain lesions compared to 2D-T2SE. Specifically, 3D-DIR showed the highest detection of intracortical and mixed WM-GM lesions, whereas 3D-FLAIR showed the highest total number of WM lesions. Both 3D-DIR and 3D-FLAIR showed the highest number of infratentorial lesions. 3D-T2 and 3D-MPRAGE did not improve lesion detection compared to 2D-T2SE. Multi-contrast, isotropic, single-slab 3D MRI allowed an improved detection of both GM and WM lesions compared to 2D-T2SE. A selection of single-slab 3D contrasts, for example, 3D-FLAIR and 3D-DIR, could replace 2D sequences in the radiological practice.
19
Alanee, Shaheen Riadh, Musatafa Deebajah, Zade Roumayah, Ali Dabaja, James O. Peabody, and Mani Menon. "Detection of significant prostate cancer through magnetic resonance imaging targeted biopsy of PI-RADS3 lesions in African American patients based on prostate specific antigen density threshold of 0.15 ng/ml2: Analysis of patient population from the Vattikuti Urology Institute." Journal of Clinical Oncology 38, no. 6_suppl (February 20, 2020): 286. http://dx.doi.org/10.1200/jco.2020.38.6_suppl.286.
Full textAbstract:
286 Background: A prostate specific antigen density (PSAD) threshold of 0.15 ng/ml2 have been suggested for significant cancer detection in PI-RADS 3 lesions to avoid unnecessary magnetic resonance imaging targeted biopsy (MRI-TB) of these lesions. However, the performance of this threshold in African American (AA) patients is not well characterized. Methods: We analyzed our institutional data base of MRI-TB to identify the rate of significant prostate cancer (Pca) detection in PI-RADS3 lesions in AA patients stratified by PSAD threshold of < 0.15 vs. ≥0.15 ng/ml2 and lesion size of < 1 cm vs ≥ 1 cm. Significant prostate cancer was defined as Gleason grade group 2 or higher on MRI-TB of the PI-RADS 3 lesion. Results: Of 768 patients included in the database, 211 (27.5%) patients identified themselves as AAs. Mean age of AA patients was 63 years and mean PSAD was 0.21. Sixty nine (32.7%) AA patients were found to have PI-RADS 3 lesions. Mean PSAD of AA patients with PI-RADS 3 lesions was 0.21 ng/ml2 as well. Fifty percent of AA patients with PI-RADS 3 lesions had PSAD ≥0.15 ng/ml2. Significant Pca detection rate for AA patients with PI-RADS 3 lesions was 9% for PSAD of ≥ 0.15 vs. 0.03% percent for AA patients with PSAD < 0.15 ng/ml2 (OR 7.056, CI 1.017-167.9, P=0.04). Stratification by lesion size (< 1 cm vs. > 1 cm) resulted in missing 0% percentage of significant Pca when only AA patients with PSAD ≥ 0.15 ng/ml2 and lesion size ≥ 1 cm received MRI-TB. Conclusions: We report on the performance of a reported PSAD density threshold in detecting significant Pca in one of the largest series of AA patients receiving MRI-TB of the prostate. Our results have direct clinical implications when counseling AA patients with PI-RADS 3 lesion on whether they should undergo MRI-TB of such lesions.
20
Gao, Junbo, Qilin Xiong, Chang Yu, and Guoqiang Qu. "White-Light Endoscopic Colorectal Lesion Detection Based on Improved YOLOv5." Computational and Mathematical Methods in Medicine 2022 (January 22, 2022): 1–11. http://dx.doi.org/10.1155/2022/9508004.
Full textAbstract:
As an effective tool for colorectal lesion detection, it is still difficult to avoid the phenomenon of missed and false detection when using white-light endoscopy. In order to improve the lesion detection rate of colorectal cancer patients, this paper proposes a real-time lesion diagnosis model (YOLOv5x-CG) based on YOLOv5 improvement. In this diagnostic model, colorectal lesions were subdivided into three categories: micropolyps, adenomas, and cancer. In the course of convolutional network training, Mosaic data enhancement strategy was used to improve the detection rate of small target polyps. At the same time, coordinate attention (CA) mechanism was introduced to take into account channel and location information in the network, so as to realize the effective extraction of three kinds of pathological features. The Ghost module was also used to generate more feature maps through linear processing, which reduces the stress of learning model parameters and speeds up detection. The experimental results show that the lesion diagnosis model proposed in this paper has a more rapid and accurate lesion detection ability, and the AP value of polyps, adenomas, and cancer is 0.923, 0.955, and 0.87, and mAP@50 is 0.916.
21
Moro, Bruna L. P., Tatiane F. Novaes, Laura R. A. Pontes, Thais Gimenez, Juan S. Lara, Daniela P. Raggio, Mariana M. Braga, and Fausto M. Mendes. "The Influence of Cognitive Bias on Caries Lesion Detection in Preschool Children." Caries Research 52, no. 5 (2018): 420–28. http://dx.doi.org/10.1159/000485807.
Full textAbstract:
We aimed to evaluate whether children’s caries experience exerts an influence on the performance of visual and radiographic methods in detecting nonevident proximal caries lesions in primary molars. Eighty children (3–6 years old) were selected and classified as having a lower (≤3 decayed, missing, or filled surfaces; dmf-s) or higher (> 3 dmf-s) caries experience. Two calibrated examiners then assessed 526 proximal surfaces for caries lesions using visual and radiographic methods. As a reference standard, 2 other examiners checked the surfaces after temporary separation. Noncavitated and cavitated lesion thresholds were considered and Poisson multilevel regression analyses were conducted to evaluate the influence of caries experience on the performance of diagnostic strategies. Accuracy parameters stratified by caries experience were also derived. A statistically significant influence of caries experience was observed only for visual inspection, with more false-positive results in children with a higher caries experience at the noncavitated lesion threshold, and more false results at the cavitated threshold. The detection of noncavitated caries lesions in children with a higher caries experience was overestimated (specificity = 0.696), compared to children with a lower caries experience (specificity = 0.918), probably due to confirmation bias. However, the examiners underestimated the detection of cavitated lesions in children with a higher caries experience (sensitivity = 0.143) compared to lower-caries-experience children (sensitivity = 0.222), possibly because of representativeness bias. The radiographic method was not influenced by children’s caries experience. In conclusion, children’s caries experience influences the performance of visual inspection in detecting proximal caries lesions in primary teeth, evidencing the occurrence of cognitive biases.
22
McCready, V. Ralph, Sabina Dizdarevic, and Thomas Beyer. "Lesion Detection and Administered Activity." Journal of Nuclear Medicine 61, no. 9 (April 3, 2020): 1406–10. http://dx.doi.org/10.2967/jnumed.120.244020.
Full text
23
Eckstein, Miguel P., and James S. Whiting. "Lesion detection in structured noise." Academic Radiology 2, no. 3 (March 1995): 249–53. http://dx.doi.org/10.1016/s1076-6332(05)80174-6.
Full text
24
Farin, P. U., Erkki Kaukanen, Heikki Jaroma, Arvi Harju, and Urho Väätäinen. "Hill-Sachs lesion: sonographic detection." Skeletal Radiology 25, no. 6 (August 1, 1996): 559–62. http://dx.doi.org/10.1007/s002560050135.
Full text
25
Williams, P. M., I. R. Francis, P. H. Bland, and C. R. Meyer. "2. LIVER-LESION BOUNDARY DETECTION." Investigative Radiology 22, no. 9 (September 1987): S1. http://dx.doi.org/10.1097/00004424-198709000-00019.
Full text
26
Ku, Yiji, Hui Ding, and Guangzhi Wang. "Efficient Synchronous Real-Time CADe for Multicategory Lesions in Gastroscopy by Using Multiclass Detection Model." BioMed Research International 2022 (August 31, 2022): 1–9. http://dx.doi.org/10.1155/2022/8504149.
Full textAbstract:
Often more than one category of lesions in patients’ gastrointestinal tracts need to be found in the endoscopic examination. Therefore, there is a need to establish an efficient synchronous real-time computer-aided detection (CADe) system for multicategory lesion detection. This paper proposes to build a system with a multiclass detection model based on the YOLOv5 to detect multicategory lesions synchronously in real-time. Two joint detection CADe systems using multiple single-class detection models with the same structure in parallel or series are established for comparison. A retrospective dataset containing 31117 images from 3747 patients is used in this study. To train the model, various online data augmentation methods and multiple loss functions are used. The proposed CADe system can synchronously detect cancers, gastrointestinal stromal tumours, polyps, and ulcers from different quality input images with 98% precision, 89% recall, and 90.2% mAP. The detection speed is 47 frames per second with a 0.04 s latency on a PC workstation. Compared to the two joint detection CADe systems, the proposed system is more accurate with faster speed and lower latency. Two extra experiments indicated that the lesion detection model based on YOLOv5x could provide better performance than other common YOLO structures and that different accuracy metrics and lesion categories have different requirements for the number of training images. The proposed synchronous real-time CADe system with the multiclass detection model can detect multicategory lesions with high accuracy and speed and low latency on limited hardware. It expands the clinical application of CADe in endoscopy and uses expensive labelled medical images more efficiently than multiple single-category lesion models for joint detection.
27
Zhang, Shujun, Shuhao Xu, Liwei Tan, Hongyan Wang, and Jianli Meng. "Stroke Lesion Detection and Analysis in MRI Images Based on Deep Learning." Journal of Healthcare Engineering 2021 (April 9, 2021): 1–9. http://dx.doi.org/10.1155/2021/5524769.
Full textAbstract:
Stroke is a kind of cerebrovascular disease that heavily damages people’s life and health. The quantitative analysis of brain MRI images plays an important role in the diagnosis and treatment of stroke. Deep neural networks with massive data learning ability supply a powerful tool for lesion detection. In order to study the property of the stroke lesions and complete intelligent automatic detection, we collaborated with two authoritative hospitals and collected 5,668 brain MRI images of 300 ischemic stroke patients. All the lesion regions in the images were accurately labeled by professional doctors to ensure the authority and effectiveness of the data. Three categories of deep learning object detection networks including Faster R-CNN, YOLOV3, and SSD are applied to implement automatic lesion detection with the best precision of 89.77%. Meanwhile, statistical analysis of the locations, shapes of the lesions, and possible related diseases is conducted with valid conclusions. The research contributes to the intelligent assisted diagnosis and prevention and treatment of ischemic stroke.
28
Ekstrand, K. R., S. Martignon, D. J. N. Ricketts, and V. Qvist. "Detection and Activity Assessment of Primary Coronal Caries Lesions: A Methodologic Study." Operative Dentistry 32, no. 3 (May 1, 2007): 225–35. http://dx.doi.org/10.2341/06-63.
Full textAbstract:
Clinical Relevance The results from this study indicate that it is possible to predict lesion depth and assess the activity of primary coronal caries lesions accurately by using the combined knowledge obtained from the visual appearance, location of the lesion and tactile sensation during probing.
29
Ching, C. Sok. "MeMoSa: Mobile Mouth Screening Anywhere for Early Detection of Oral Cancer." Journal of Global Oncology 4, Supplement 2 (October 1, 2018): 56s. http://dx.doi.org/10.1200/jgo.18.33300.
Full textAbstract:
Background: The burden of oral cancer is significant in Asia, accounting for 11% of cancer deaths in Asia compared with 5% in the rest of the world. High and increasing access to mobile phones in low- and middle-income countries presents an opportunity to use telemedicine to facilitate early detection of oral cancer. Aim: To evaluate the feasibility of using the mobile phone as a documentation and communication tool for early detection of high-risk oral lesions. Methods: We determined the concordance between evaluation of oral lesions using mobile phone-captured images and clinical oral examination using kappa statistics. A mobile phone app named MeMoSA was developed and the app was tested during a routine oral cancer screening program in the community to determine the feasibility of integrating this tool for the documentation of oral lesions, and the communication between dentists and specialists with regards to management of these patients. The experience of dentists and specialists in using MeMoSA was determined using qualitative questionnaires. Results: We demonstrated that the mobile phone is a sensitive and specific tool, with a sensitivity of > 80% in detecting a lesion, an accuracy of 87% in categorizing the type of lesion and 85% concordance in patient referral compared with clinical oral examination. Having been trained to use MeMoSA, 36/36 dentists agreed that this app could improve early detection of oral mucosal lesions. All of these dentists wanted to continue using the app as screening tool in the future as they believe that it could assist them in the identification of high-risk oral mucosal lesions through direct communication with specialists. Conclusion: MeMoSA enabled documentation of the lesion through easy photography and facilitated patient management through quick communication between dentists and specialists. Because of its ease of use MeMoSA could be useful tool in early detection of high-risk oral lesions in low-resource settings and could increase the access to healthcare in geographically hard to reach populations.
30
Li, Meichen, Delan Li, Xue Hou, Xiangheng Zhang, Na Wang, Jianzhong Liang, Jing Chen, et al. "Utilizing phenotypic characteristics of metastatic brain tumors to predict the probability of circulating tumor DNA detection from cerebrospinal fluid." Journal of Clinical Oncology 38, no. 15_suppl (May 20, 2020): 2507. http://dx.doi.org/10.1200/jco.2020.38.15_suppl.2507.
Full textAbstract:
2507 Background: Brain metastases occur in approximately 20% of tumor patients and is often associated with terminal events and poor prognosis. Cerebrospinal fluid (CSF) can be a promising source for detecting circulating tumor DNA (ctDNA) specific to the central nervous system (CNS) instead of peripheral blood due to the blood-brain barrier. However, CSF’s suboptimal ctDNA detection rate might limit its clinical application. Precise screening of suitable patients is needed to maximize clinical benefit. Methods: We sequenced 425 cancer-relevant genes in CSF and matched extracranial tissue or blood samples obtained from 67 lung cancer patients with brain metastases. The impact of clinical factors, including age, gender, tumor size, number of lesions, and distance of lesions to the ventricle on CSF ctDNA detection was then evaluated by univariate logistic regression. To predict the probability of successful CSF ctDNA detection, best subsets regression was employed for feature selection and cross validation was used for performance assessment to determine the final model. Results: We detected somatic alterations in 39/67 (58%) CSF ctDNA, 57/66 (86%) plasma ctDNA and 45/49 (92%) tissue samples. Mutation detection rate of CSF ctDNA was significantly lower than that from extracranial tissue and plasma (P < 0.001). Univariate analysis revealed significant association (P < 0.05) of high CSF ctDNA detection rate with the following features: (1) intracranial lesion size ( T), (2) shortest distance between the largest lesion and the ventricle ( Dtop), and (3) shortest distance between all intracranial lesion and the ventricle ( Dall). We also revealed a trend of higher detection rate in patients with CNS symptoms ( SCNS). Subsequent best subsets analysis and cross validation suggested best prediction power with lesion size and largest lesion-ventriclar distance (area under curve [AUC], 0.76 [95% CI, 0.71 to 0.85]; accuracy, 0.75 [95% CI, 0.70 to 0.81]). Final probability can then be derived from Logit P = 0.11×T−0.16×Dall (AUC, 0.82; sensitivity, 0.91; specificity, 0.74). The detection of CSF ctDNA was significantly improved from 58% to 83% (P = 0.03) based on the model. Conclusions: This study established a regression model to predict the probability of CSF ctDNA that can be useful to facilitate clinical decisions and avoid excessive practice when monitoring tumor evolution in the brain.
31
AlShaya, Mohammad S., Heba J. Sabbagh, and Azza A. El-Housseiny. "Diagnosis and Management Approaches for Non-cavitated Carious Dental Lesions- A Narrative Review." Open Dentistry Journal 15, no. 1 (August 24, 2021): 337–47. http://dx.doi.org/10.2174/1874210602115010337.
Full textAbstract:
Background: Dental caries is one of the most prevalent diseases. Its detection and management should start with a comprehensive treatment plan, with the goals of the elimination of cariogenic bacteria, the reduction of plaque acidogenicity, the encouragement of tooth remineralization, and the repair of damaged teeth. Objectives: The aim of this paper was to review the literature regarding the latest updates on the diagnosis and management approaches of non-cavitated carious dental lesions. Methods: Studies regarding the diagnosis and management of non-cavitated carious dental lesions were included. Results: The subclinical non-cavitated carious lesion might progress to an early enamel lesion, develop into an established dentin lesion, or sometimes end up with a lesion reaching the pulp. The detection and management of caries should be patient-centered, risk-based, and evidence-supported, and should consider the dentists’ expertise and the patients’ needs and preferences. The visual-tactile and radiographic detection of non-cavitated carious lesions are greatly helped by the advances of non-invasive detection tools such as DIAGNOdent, fiber-optic transillumination, quantitative light-induced fluorescence, and DIAGNOcam. Conclusion: Accordingly, non-cavitated carious lesions can be arrested by several non-invasive techniques, which are preferred over the invasive options. The clinicians can use sealants plus fluoride varnish on occlusal surfaces, fluoride varnish or resin infiltration on proximal surfaces, and resin infiltration,fluoride gel, or varnish alone on facial or lingual surfaces to manage non-cavitated carious lesions.
32
Nyvad, Bente, and Vibeke Baelum. "Nyvad Criteria for Caries Lesion Activity and Severity Assessment: A Validated Approach for Clinical Management and Research." Caries Research 52, no. 5 (2018): 397–405. http://dx.doi.org/10.1159/000480522.
Full textAbstract:
The Nyvad classification is a visual-tactile caries classification system devised to enable the detection of the activity and severity of caries lesions with special focus on low-caries populations. The criteria behind the classification reflect the entire continuum of caries, ranging from clinically sound surfaces through noncavitated and microcavitated caries lesions in enamel, to frank cavitation into the dentin. Lesion activity at each severity stage is discriminated by differences in surface topography and lesion texture. The reliability of the Nyvad criteria is high to excellent when used by trained examiners in the primary and permanent dentitions. The Nyvad criteria have construct validity for lesion activity assessments because of their ability to reflect the well-known caries-controlling effect of fluoride. Predictive validity was demonstrated by showing that active noncavitated lesions are at higher risk of progressing to a cavity or filled state than do inactive noncavitated lesions. Lesion activity assessment performed successfully as a screening tool to identify individuals with a poor caries prognosis. Because of their predictive validity, the Nyvad criteria are superior to other current caries lesion descriptors for the detection of changes in the lesion activity status over time. The Nyvad criteria fulfill all the formal requirements for a robust caries lesion classification and are recommended for evidence-based caries management in clinical practice and in research.
33
Goujon, Adrien, Sonia Mirafzal, Kevin Zuber, Romain Deschamps, Jean-Claude Sadik, Olivier Gout, Julien Savatovsky, and Augustin Lecler. "3D-Fast Gray Matter Acquisition with Phase Sensitive Inversion Recovery Magnetic Resonance Imaging at 3 Tesla: Application for detection of spinal cord lesions in patients with multiple sclerosis." PLOS ONE 16, no. 4 (April 22, 2021): e0247813. http://dx.doi.org/10.1371/journal.pone.0247813.
Full textAbstract:
Background and purpose To compare 3D-Fast Gray Matter Acquisition with Phase Sensitive Inversion Recovery (3D-FGAPSIR) with conventional 3D-Short-Tau Inversion Recovery (3D-STIR) and sagittal T1-and T2-weighted MRI dataset at 3 Tesla when detecting MS spinal cord lesions. Material and methods This prospective single-center study was approved by an institutional review board and enrolled participants from December 2016 to August 2018. Two neuroradiologists blinded to all data, individually analyzed the 3D-FGAPSIR and the conventional datasets separately and in random order. Discrepancies were resolved by consensus by a third neuroradiologist. The primary judgment criterion was the number of MS spinal cord lesions. Secondary judgment criteria included lesion enhancement, lesion delineation, reader-reported confidence and lesion-to-cord-contrast-ratio. A Wilcoxon’s test was used to compare the two datasets. Results 51 participants were included. 3D-FGAPSIR detected significantly more lesions than the conventional dataset (344 versus 171 respectively, p<0.001). Two participants had no detected lesion on the conventional dataset, whereas 3D-FGAPSIR detected at least one lesion. 3/51 participants had a single enhancing lesion detected by both datasets. Lesion delineation and reader-reported confidence were significantly higher with 3D-FGAPSIR: 4.5 (IQR 1) versus 2 (IQR 0.5), p<0.0001 and 4.5 (IQR 1) versus 2.5 (IQR 0.5), p<0.0001. Lesion-to-cord-contrast-ratio was significantly higher using 3D-FGAPSIR as opposed to 3D-STIR and T2: 1.4 (IQR 0,3) versus 0.4 (IQR 0,1) and 0.3 (IQR 0,1)(p = 0.04). Correlations with clinical data and inter- and intra-observer agreements were higher with 3D-FGAPSIR. Conclusion 3D-FGAPSIR improved overall MS spinal cord lesion detection as compared to conventional set and detected all enhancing lesions.
34
Xie, Yilin, Zhuoyue Wu, Xin Han, Hongyu Wang, Yifan Wu, Lei Cui, Jun Feng, Zhaohui Zhu, and Zhongyuanlong Chen. "Computer-Aided System for the Detection of Multicategory Pulmonary Tuberculosis in Radiographs." Journal of Healthcare Engineering 2020 (August 24, 2020): 1–12. http://dx.doi.org/10.1155/2020/9205082.
Full textAbstract:
The early screening and diagnosis of tuberculosis plays an important role in the control and treatment of tuberculosis infections. In this paper, an integrated computer-aided system based on deep learning is proposed for the detection of multiple categories of tuberculosis lesions in chest radiographs. In this system, the fully convolutional neural network method is used to segment the lung area from the entire chest radiograph for pulmonary tuberculosis detection. Different from the previous analysis of the whole chest radiograph, we focus on the specific tuberculosis lesion areas for the analysis and propose the first multicategory tuberculosis lesion detection method. In it, a learning scalable pyramid structure is introduced into the Faster Region-based Convolutional Network (Faster RCNN), which effectively improves the detection of small-area lesions, mines indistinguishable samples during the training process, and uses reinforcement learning to reduce the detection of false-positive lesions. To compare our method with the current tuberculosis detection system, we propose a classification rule for whole chest X-rays using a multicategory tuberculosis lesion detection model and achieve good performance on two public datasets (Montgomery: AUC = 0.977 and accuracy = 0.926; Shenzhen: AUC = 0.941 and accuracy = 0.902). Our proposed computer-aided system is superior to current systems that can be used to assist radiologists in diagnoses and public health providers in screening for tuberculosis in areas where tuberculosis is endemic.
35
Ahmed, Iman Emad, Hayder Kareem Al-Jaberi, and Mohammed M. Jawad Alkahlissi. "Comparison of proton density MRI and T2-Weighted Fast Echo for The Detection of Cervical Spinal Cord Multiple Sclerosis Lesions." Journal of the Faculty of Medicine Baghdad 60, no. 4 (May 12, 2019): 195–201. http://dx.doi.org/10.32007/jfacmedbagdad.604159.
Full textAbstract:
Background: The prevalence of spinal cord lesions is high in multiple sclerosis particularly in the cervical cord, and their detection can assist in both the diagnosis and follow-up of the patients. For spinal multiple sclerosis, MRI is considered the first line investigation.
Objective: To evaluate the value of sagittal 1.5 Tesla proton density-fast spin echo (PD-FSE) MRI in the detecting and increasing conspicuity of multiple sclerosis lesions in cervical cord in comparison with sagittal T2 fast spin-echo (T2-FSE) MRI.
Patients and Methods: A cross sectional study carried out from 3rd of January 2017 to 1st of January 2018 in the MRI department of Al-Imamein Al-Kadhimein Medical City, and included 60 selected patients with a known diagnosis of multiple sclerosis. All patients were examined with 1.5 T sagittal PD-FSE, T2-FSE and axial gradient recalled-echo (GRE) MRI.
Results: Sixty patients with cervical multiple sclerosis were enrolled in the study, 146 (100%) lesions were detected by PD-FSE imaging, while T2 detected 105 (71.9%), 41 more lesions (28%) were detected by PD-FSE imaging, (P-value <0.001). All extra lesions were confirmed on axial imaging. In 13 patients (21.6%) one lesion or more had been detected on sagittal PD-FSE imaging while on sagittal T2-FSE imaging, no lesion were detected. On PD-FSE imaging, 17 long lesions were detected in 16 patients (26.7%) while 7 long lesions in 7 patients (11.7%) were detected by T2-FSE imaging. So, in 9 patients (16.7%) 10 lesions were detected as long in PD-FSE while short lesion in T2– FSE, the detection of long lesions by PD-FSE was significantly higher than in T2– FSE (100% vs 71.9% with p- value of 0.002). The mean lesion contrast to cord ratio was significantly higher in PD-FSE as compared to T2-FSE (PD-FSE, 79±2.0, against T2-FSE, 61± 2.6; P-value <0.001).
Conclusion: Sagittal proton density was more efficient and more accurate in the detection of cervical cord lesions than sagittal T2-FSE sequence, when used in conjunction with sagittal T2-FSE; it can raise the diagnostic assurance via improving the visualization of the lesions.
36
Lu, Amanda, Kamyar Ghabili, Kevin Nguyen, and Preston Sprenkle. "Role of core location in targeted MRI-ultrasound fusion biopsy of prostate lesions." Journal of Clinical Oncology 36, no. 6_suppl (February 20, 2018): 136. http://dx.doi.org/10.1200/jco.2018.36.6_suppl.136.
Full textAbstract:
136 Background: Targeted mpMRI fusion biopsy has gained adoption with superior clinically significant cancer detection rates and accuracy over template biopsy. We sought to establish the role of biopsy location within a prostate lesion to detect clinically significant prostate cancer. Methods: From Nov 2016-Aug 2017, 110 patients with positive multiparametric-MRI (mpMRI) underwent targeted and systematic MRI-US fusion biopsy at our institution for clinical suspicion or known history of prostate cancer. Lesions were scored by Prostate imaging reporting and data system (PI-RADS) classification schema by experienced genitourinary radiologists. Biopsy was performed by an oncology-trained urologist (PS) performing a high volume of fusion biopsies. 5 cores were taken from each lesion, each corresponding to a predetermined location (central, medial, lateral, apex, and base of the lesion). Cancer detection rates (CDR) were calculated on a per lesion basis from biopsy histology. Results: 154 prostate lesions were identified and biopsied with an average volume of 1.31 mL. Detection of clinically significant cancer (G>3+4) did not differ significantly among the 5 locations (Table 1). The central core detected slightly more G≥3+4 cancers than the apex core. No concordance of pathology grade was found between the central core and location of the peripheral core (medial, lateral, apex, or base). In 32% (50/154) of lesions, the peripheral cores had a higher Gleason score than the central core. Biopsy of only the central core missed 40% (21/52) of G≥3+4 cancers and 17% (4/24) of G>3+4 cancers. Lesions with higher PIRADs score were more likely to detect cancer in both the central and peripheral cores, but lesion volume was not a significant predictor. Conclusions: Location of biopsy cores within mpMRI-identified prostate lesions has little correlation with detection of clinically significant cancer. However, targeted biopsy of only the center of a lesion can miss 17% of Gleason >3+4 cancers. [Table: see text]
37
Granata, Vincenza, Roberta Fusco, Antonio Avallone, Antonino Cassata, Raffaele Palaia, Paolo Delrio, Roberta Grassi, et al. "Abbreviated MRI protocol for colorectal liver metastases: How the radiologist could work in pre surgical setting." PLOS ONE 15, no. 11 (November 19, 2020): e0241431. http://dx.doi.org/10.1371/journal.pone.0241431.
Full textAbstract:
Background MRI is the most reliable imaging modality that allows to assess liver metastases. Our purpose is to compare the per-lesion and per-patient detection rate of gadoxetic acid-(Gd-EOB) enhanced liver MRI and fast MR protocol including Diffusion Weighted Imaging (DWI) and T2-W Fat Suppression sequence in the detection of liver metastasis in pre surgical setting. Methods One hundred and eight patients with pathologically proven liver metastases (756 liver metastases) underwent Gd-EOBMRI were enrolled in this study. Three radiologist independently graded the presence of liver lesions on a five-point confidence scale assessed only abbreviated protocol (DWI and sampling perfection with application-optimized contrasts using different flip angle evolution (SPACE) fat suppressed sequence) and after an interval of more than 2 weeks the conventional study (all acquired sequences). Per-lesion and per-patient detection rate of metastases were calculated. Weighted к values were used to evaluate inter-reader agreement of the confidence scale regarding the presence of the lesion. Results MRI detected 732 liver metastases. All lesions were identified both by conventional study as by abbreviated protocol. In terms of per-lesion detection rate of liver metastasis, all three readers had higher detection rate both with abbreviated protocol and with standard protocol with Gd-EOB (96.8% [732 of 756] vs. 96.5% [730 of 756] for reader 1; 95.8% [725 of 756] vs. 95.2% [720 of 756] for reader 2; 96.5% [730 of 756] vs. 96.5% [730 of 756] for reader 3). Inter-reader agreement of lesions detection rate between the three radiologists was excellent (k range, 0.86–0.98) both for Gd-EOB MRI and for Fast protocol (k range, 0.89–0.99). Conclusion Abbreviated protocol showed the same detection rate than conventional study in detection of liver metastases.
38
Xu, Guang, Minghua Yao, Jian Wu, Lehang Guo, Lijing Feng, Shuai Wang, Lixia Zhao, Huixiong Xu, and Rong Wu. "Diagnostic Value of Different Systematic Prostate Biopsy Methods in the Detection of Prostate Cancer with Ultrasonographic Hypoechoic Lesions - A Comparative Study." Urologia Internationalis 95, no. 2 (2015): 183–88. http://dx.doi.org/10.1159/000381752.
Full textAbstract:
Objective: To assess if a less extended biopsy in the transperineal approach is sufficient for detection of prostate cancer (PC) in patients with hypoechoic lesions. Methods: This was a prospective study of 167 consecutive patients with prostate hypoechoic lesion and who underwent transperineal ultrasound (TPUS)-guided 12-core and hypoechoic lesion core biopsy between January 2012 and February 2013. Results: PC was detected in 64.1% (107/167) of patients. The PC detection rate of the 12-core prostate biopsy scheme was the highest, but when including the hypoechoic lesion core, there was no difference between the 6- and 12-core schemes (all p > 0.05), irrespective of prostate volume or prostate-specific antigen levels (all p > 0.05). Conclusions: A more limited biopsy scheme could be sufficient for the detection of PC if the hypoechoic lesion is sampled.
39
Guaragnella, Cataldo, and Maria Rizzi. "Simple and Accurate Border Detection Algorithm for Melanoma Computer Aided Diagnosis." Diagnostics 10, no. 6 (June 22, 2020): 423. http://dx.doi.org/10.3390/diagnostics10060423.
Full textAbstract:
The interest of the scientific community for computer aided skin lesion analysis and characterization has been increased during the last years for the growing incidence of melanoma among cancerous pathologies. The detection of melanoma in its early stage is essential for prognosis improvement and for guaranteeing a high five-year relative survival rate of patients. The clinical diagnosis of skin lesions is challenging and not trivial since it depends on human vision and physician experience and expertise. Therefore, a computer method that makes an accurate extraction of important details of skin lesion image can assist dermatologists in cancer detection. In particular, the border detection is a critical computer vision issue owing to the wide range of lesion shapes, sizes, colours and skin texture types. In this paper, an automatic and effective pigmented skin lesion segmentation method in dermoscopic image is presented. The proposed procedure is adopted to extract a mask of the lesion region without the adoption of other signal processing procedures for image improvement. A quantitative experimental evaluation has been performed on a publicly available database. The achieved results show the method validity and its high robustness towards irregular boundaries, smooth transition between lesion and skin, noise and artifact presence.
40
Polonara, G., and U. Salvolini. "Turbo-FLAIR Sequence in Brain MRI." Rivista di Neuroradiologia 11, no. 1 (February 1998): 27–37. http://dx.doi.org/10.1177/197140099801100103.
Full textAbstract:
To compare a turbo-FLAIR sequence with Proton Density (PD) and T2-weighted Turbo Spin-Echo (TSE) sequences in several different brain diseases, 276 MRI examinations were performed on a 1.0 Tesla system. The positive cases were assessed for lesion detection and lesion conspicuity. Four quantitative criteria were also used to compare the contrast of the two techniques: lesion to background contrast, lesion to background contrast to noise ratio (CNR), lesion to Cerobrospinal fluid (CSF) contrast, lesion to CSF contrast to noise ratio. The turbo FLAIR sequence was found to be superior to PD and T2-weighted TSE for lesion detection: this sequence detected more lesions in 74 patients than PD and in 42 patients than T2, but missed some subtentorial lesions. For lesion conspicuity turbo-FLAIR was judged equivalent to PD and T2-weighted TSE respectively in 27% and 45% of the cases and better in 71% and 53% of the cases. Lesion to background contrast and lesion to background CNR were found to be significantly greater for turbo-FLAIR than for PD (p<0.001). Compared with T2-weighted TSE, turbo-FLAIR showed a significantly higher lesion to background contrast (p<0.001) and inferior lesion to background CNR (p<0.001). Our study indicates that turbo-FLAIR can replace PD TSE scans in most cases and can be used as a first choice sequence for cerebrovascular diseases, multiple sclerosis and for the evaluation of gliosis.
41
Courtenay, Lloyd A., Inés Barbero-García, Julia Aramendi, Diego González-Aguilera, Manuel Rodríguez-Martín, Pablo Rodríguez-Gonzalvez, Javier Cañueto, and Concepción Román-Curto. "A Novel Approach for the Shape Characterisation of Non-Melanoma Skin Lesions Using Elliptic Fourier Analyses and Clinical Images." Journal of Clinical Medicine 11, no. 15 (July 28, 2022): 4392. http://dx.doi.org/10.3390/jcm11154392.
Full textAbstract:
The early detection of Non-Melanoma Skin Cancer (NMSC) is crucial to achieve the best treatment outcomes. Shape is considered one of the main parameters taken for the detection of some types of skin cancer such as melanoma. For NMSC, the importance of shape as a visual detection parameter is not well-studied. A dataset of 993 standard camera images containing different types of NMSC and benign skin lesions was analysed. For each image, the lesion boundaries were extracted. After an alignment and scaling, Elliptic Fourier Analysis (EFA) coefficients were calculated for the boundary of each lesion. The asymmetry of lesions was also calculated. Then, multivariate statistics were employed for dimensionality reduction and finally computational learning classification was employed to evaluate the separability of the classes. The separation between malignant and benign samples was successful in most cases. The best-performing approach was the combination of EFA coefficients and asymmetry. The combination of EFA and asymmetry resulted in a balanced accuracy of 0.786 and an Area Under Curve of 0.735. The combination of EFA and asymmetry for lesion classification resulted in notable success rates when distinguishing between benign and malignant lesions. In light of these results, skin lesions’ shape should be integrated as a fundamental part of future detection techniques in clinical screening.
42
Ali, Abder-Rahman, Jingpeng Li, Guang Yang, and Sally Jane O’Shea. "A machine learning approach to automatic detection of irregularity in skin lesion border using dermoscopic images." PeerJ Computer Science 6 (June 29, 2020): e268. http://dx.doi.org/10.7717/peerj-cs.268.
Full textAbstract:
Skin lesion border irregularity is considered an important clinical feature for the early diagnosis of melanoma, representing the B feature in the ABCD rule. In this article we propose an automated approach for skin lesion border irregularity detection. The approach involves extracting the skin lesion from the image, detecting the skin lesion border, measuring the border irregularity, training a Convolutional Neural Network and Gaussian naive Bayes ensemble, to the automatic detection of border irregularity, which results in an objective decision on whether the skin lesion border is considered regular or irregular. The approach achieves outstanding results, obtaining an accuracy, sensitivity, specificity, and F-score of 93.6%, 100%, 92.5% and 96.1%, respectively.
43
RAJU K, RAINA, and S. Swapna Kumar. "A Comparative Study of Various Techniques used for Melanoma Detection." International Journal of Advanced Research in Computer Science and Software Engineering 7, no. 11 (November 30, 2017): 44. http://dx.doi.org/10.23956/ijarcsse.v7i11.466.
Full textAbstract:
Skin cancer is one of the most fatal disease. It is easily curable, when it is detected in its beginning stage. Early detection of melanoma through accurate techniques and innovative technologies has the greatest potential for decreasing mortality associated with this disease. Mainly there are four steps for detecting melanoma which includes preprocessing, segmentation, feature extraction and classification. The preprocessing stage will remove all the artifacts associated with the lesion. The exact boundaries of lesion are identified from normal skin through segmentation method. Feature extraction stage is used for calculating and obtaining different parameters of the lesion region. The final stage is to classify the lesion as benign or malignant. In this paper different types of segmentation methods and classification methods are described. Both of these stages are accurately implemented to reach the final detection of the lesion.
44
K, Raina Raju, and S. Swapna Kumar. "A Comparative Study of Various Techniques used for Melanoma Detection." International Journal of Advanced Research in Computer Science and Software Engineering 7, no. 9 (October 31, 2017): 52. http://dx.doi.org/10.23956/ijarcsse.v7i9.411.
Full textAbstract:
Skin cancer is one of the most fatal disease. It is easily curable, when it is detected in its beginning stage. Early detection of melanoma through accurate techniques and innovative technologies has the greatest potential for decreasing mortality associated with this disease. Mainly there are four steps for detecting melanoma which includes preprocessing, segmentation, feature extraction and classification. The preprocessing stage will remove all the artifacts associated with the lesion. The exact boundaries of lesion are identified from normal skin through segmentation method. Feature extraction stage is used for calculating and obtaining different parameters of the lesion region. The final stage is to classify the lesion as benign or malignant. In this paper different types of segmentation methods and classification methods are described. Both of these stages are accurately implemented to reach the final detection of the lesion.
45
Jha, A., and AK Chaurasia. "A retrospective analysis of cervical smears for detection of precancerous lesions." Journal of Pathology of Nepal 5, no. 10 (September 14, 2015): 847–49. http://dx.doi.org/10.3126/jpn.v5i10.15641.
Full textAbstract:
Background: Diseases of the cervix are common in women. Detection of neoplastic lesions is of prime importance in the evaluation of cervical smears although their frequency is less than the non-neoplastic conditions. Reactive and inflammatory conditions, however, may mimic or obscure the dysplastic changes. The aim of this study was to evaluate the frequencies of various inflammatory, reactive and neoplastic lesions in the cervical smears.Materials and Methods: This was a retrospective study of 150 cases of cervical smears collected from medical record section of the hospital and Department of Pathology at National Medical College from April 2013 to April 2015.Results: Of 150 patients who underwent cervical cytology, only 15 cases (10%) had epithelial cell abnormalities. This was followed by reactive cellular changes associated with inflammation (16.7%), infections (5.3%) and atrophy (1.3%). Low grade squamous intraepithelial lesion (6%) was the most common epithelial cell abnormality followed by high grade squamous intraepithelial lesion (2%), atypical glandular cells (1.3%) and squamous cell carcinoma (0.7%).Conclusion: Reactive changes including atrophy were the commonest finding. Squamous intraepithelial lesion was commonest finding among epithelial abnormalities. Glandular intraepithelial lesions and squamous cell carcinoma were also identified. Among infections Trichomoniasis and candidiasis were seen.
46
Fisichella, V. A., F. Jäderling, S. Horvath, P. O. Stotzer, A. Kilander, and M. Hellström. "Primary three-dimensional analysis with perspective-filet view versus primary two-dimensional analysis: Evaluation of lesion detection by inexperienced readers at computed tomographic colonography in symptomatic patients." Acta Radiologica 50, no. 3 (April 2009): 244–55. http://dx.doi.org/10.1080/02841850802714797.
Full textAbstract:
Background: “Perspective-filet view” is a novel three-dimensional (3D) viewing technique for computed tomography colonography (CTC). Studies with experienced readers have shown a sensitivity for perspective-filet view similar to that of 2D or 3D endoluminal fly-through in detection of colorectal lesions. It is not known whether perspective-filet view, compared to axial images, improves lesion detection by inexperienced readers. Purpose: To compare primary 3D analysis using perspective-filet view (3D Filet) with primary 2D analysis, as used by inexperienced CTC readers. Secondary aims were to compare lesion detection by 3D Filet when used by experienced and inexperienced readers, and to evaluate the effect of combined 3D Filet + 2D analysis. Material and Methods: Fifty symptomatic patients were prospectively enrolled. An experienced reader performed 3D Filet analysis followed by complete 2D analysis (3D Filet + 2D), before colonoscopy with segmental unblinding. Two inexperienced readers (readers 2 and 3), blinded to CTC and colonoscopy findings, retrospectively performed 3D Filet analysis and, after 5 weeks, 2D analysis. True positives ≥6 mm detected by the inexperienced readers with 3D Filet and/or 2D were combined to obtain 3D Filet + 2D. Results: Colonoscopy revealed 116 lesions: 16 lesions ≥10 mm, 19 lesions 6–9 mm, and 81 lesions ≤5 mm. For the experienced reader, sensitivities for lesions ≥6 mm with 3D Filet and 3D Filet + 2D were 77% and 83%, respectively. For the inexperienced readers, sensitivities for lesions ≥6 mm with 3D Filet and 2D were 51% and 57% (reader 2) and 40% and 43% (reader 3), respectively. There was no significant difference between 3D Filet and 2D regarding sensitivity and reading time. For lesions ≥6 mm, 3D Filet + 2D improved the sensitivity of reader 2 to 63% and of reader 3 to 51%. Conclusion: Lesion detection by inexperienced readers using perspective-filet view is comparable to that obtained by 2D. Lesion detection improves by combining 3D Filet + 2D, but not to the level of an experienced reader.
47
Oh, Dong Jun, Youngbae Hwang, and Yun Jeong Lim. "A Current and Newly Proposed Artificial Intelligence Algorithm for Reading Small Bowel Capsule Endoscopy." Diagnostics 11, no. 7 (June 29, 2021): 1183. http://dx.doi.org/10.3390/diagnostics11071183.
Full textAbstract:
Small bowel capsule endoscopy (SBCE) is one of the most useful methods for diagnosing small bowel mucosal lesions. However, it takes a long time to interpret the capsule images. To solve this problem, artificial intelligence (AI) algorithms for SBCE readings are being actively studied. In this article, we analyzed several studies that applied AI algorithms to SBCE readings, such as automatic lesion detection, automatic classification of bowel cleanliness, and automatic compartmentalization of small bowels. In addition to automatic lesion detection using AI algorithms, a new direction of AI algorithms related to shorter reading times and improved lesion detection accuracy should be considered. Therefore, it is necessary to develop an integrated AI algorithm composed of algorithms with various functions in order to be used in clinical practice.
48
Wang, Xueling, Xianmin Meng, and Shu Yan. "Deep Learning-Based Image Segmentation of Cone-Beam Computed Tomography Images for Oral Lesion Detection." Journal of Healthcare Engineering 2021 (September 21, 2021): 1–7. http://dx.doi.org/10.1155/2021/4603475.
Full textAbstract:
This paper aimed to study the adoption of deep learning (DL) algorithm of oral lesions for segmentation of cone-beam computed tomography (CBCT) images. 90 patients with oral lesions were taken as research subjects, and they were grouped into blank, control, and experimental groups, whose images were treated by the manual segmentation method, threshold segmentation algorithm, and full convolutional neural network (FCNN) DL algorithm, respectively. Then, effects of different methods on oral lesion CBCT image recognition and segmentation were analyzed. The results showed that there was no substantial difference in the number of patients with different types of oral lesions among three groups ( P > 0.05 ). The accuracy of lesion segmentation in the experimental group was as high as 98.3%, while those of the blank group and control group were 78.4% and 62.1%, respectively. The accuracy of segmentation of CBCT images in the blank group and control group was considerably inferior to the experimental group ( P < 0.05 ). The segmentation effect on the lesion and the lesion model in the experimental group and control group was evidently superior to the blank group ( P < 0.05 ). In short, the image segmentation accuracy of the FCNN DL method was better than the traditional manual segmentation and threshold segmentation algorithms. Applying the DL segmentation algorithm to CBCT images of oral lesions can accurately identify and segment the lesions.
49
Dellios, Nikolaos, Ulf Teichgraeber, Robert Chelaru, Ansgar Malich, and Ismini E. Papageorgiou. "Computer-aided Detection Fidelity of Pulmonary Nodules in Chest Radiograph." Journal of Clinical Imaging Science 7 (February 20, 2017): 8. http://dx.doi.org/10.4103/jcis.jcis_75_16.
Full textAbstract:
Aim: The most ubiquitous chest diagnostic method is the chest radiograph. A common radiographic finding, quite often incidental, is the nodular pulmonary lesion. The detection of small lesions out of complex parenchymal structure is a daily clinical challenge. In this study, we investigate the efficacy of the computer-aided detection (CAD) software package SoftView™ 2.4A for bone suppression and OnGuard™ 5.2 (Riverain Technologies, Miamisburg, OH, USA) for automated detection of pulmonary nodules in chest radiographs. Subjects and Methods: We retrospectively evaluated a dataset of 100 posteroanterior chest radiographs with pulmonary nodular lesions ranging from 5 to 85 mm. All nodules were confirmed with a consecutive computed tomography scan and histologically classified as 75% malignant. The number of detected lesions by observation in unprocessed images was compared to the number and dignity of CAD-detected lesions in bone-suppressed images (BSIs). Results: SoftView™ BSI does not affect the objective lesion-to-background contrast. OnGuard™ has a stand-alone sensitivity of 62% and specificity of 58% for nodular lesion detection in chest radiographs. The false positive rate is 0.88/image and the false negative (FN) rate is 0.35/image. From the true positive lesions, 20% were proven benign and 80% were malignant. FN lesions were 47% benign and 53% malignant. Conclusion: We conclude that CAD does not qualify for a stand-alone standard of diagnosis. The use of CAD accompanied with a critical radiological assessment of the software suggested pattern appears more realistic. Accordingly, it is essential to focus on studies assessing the quality-time-cost profile of real-time (as opposed to retrospective) CAD implementation in clinical diagnostics.
50
Lu, Feng, Wei Li, Song Lin, Chengwangli Peng, Zhiyong Wang, Bin Qian, Rajiv Ranjan, Hai Jin, and Albert Y. Zomaya. "Multi-scale Features Fusion for the Detection of Tiny Bleeding in Wireless Capsule Endoscopy Images." ACM Transactions on Internet of Things 3, no. 1 (February 28, 2022): 1–19. http://dx.doi.org/10.1145/3477540.
Full textAbstract:
Wireless capsule endoscopy is a modern non-invasive Internet of Medical Imaging Things that has been increasingly used in gastrointestinal tract examination. With about one gigabyte image data generated for a patient in each examination, automatic lesion detection is highly desirable to improve the efficiency of the diagnosis process and mitigate human errors. Despite many approaches for lesion detection have been proposed, they mainly focus on large lesions and are not directly applicable to tiny lesions due to the limitations of feature representation. As bleeding lesions are a common symptom in most serious gastrointestinal diseases, detecting tiny bleeding lesions is extremely important for early diagnosis of those diseases, which is highly relevant to the survival, treatment, and expenses of patients. In this article, a method is proposed to extract and fuse multi-scale deep features for detecting and locating both large and tiny lesions. A feature extracting network is first used as our backbone network to extract the basic features from wireless capsule endoscopy images, and then at each layer multiple regions could be identified as potential lesions. As a result, the features maps of those potential lesions are obtained at each level and fused in a top-down manner to the fully connected layer for producing final detection results. Our proposed method has been evaluated on a clinical dataset that contains 20,000 wireless capsule endoscopy images with clinical annotation. Experimental results demonstrate that our method can achieve 98.9% prediction accuracy and 93.5% score, which has a significant performance improvement of up to 31.69% and 22.12% in terms of recall rate and score, respectively, when compared to the state-of-the-art approaches for both large and tiny bleeding lesions. Moreover, our model also has the highest AP and the best medical diagnosis performance compared to state-of-the-art multi-scale models.