Deep learning in computational dermatopathology of melanoma: A technical systematic literature review

Sauter, Daniel; Lodde, Georg; Nensa, Felix; Schadendorf, Dirk; Livingstone, Elisabeth; Kukuk, Markus

doi:10.1016/j.compbiomed.2023.107083

Cited by 14 publications

(8 citation statements)

References 138 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…False negatives results were most commonly attributed to "technical" factors, across all four medical centers and regardless of the scanner model used. Factors such as differences in staining intensity, color, artifacts and scanning resolution and focus, appear to affect the DSA's ability to correctly identify a ganglion cell, to a greater degree than these same factors might affect a trained human observer [ 48 ]. Additional samples for training and validation of the DSA may further improve its performance and minimize the effects of technical factors.…”

Section: Discussionmentioning

confidence: 99%

Algorithm-assisted diagnosis of Hirschsprung’s disease – evaluation of robustness and comparative image analysis on data from various labs and slide scanners

Greenberg,

Samueli,

Farkash

et al. 2024

Diagn Pathol

View full text Add to dashboard Cite

Background Differences in the preparation, staining and scanning of digital pathology slides create significant pre-analytic variability. Algorithm-assisted tools must be able to contend with this variability in order to be applicable in clinical practice. In a previous study, a decision support algorithm was developed to assist in the diagnosis of Hirschsprung's disease. In the current study, we tested the robustness of this algorithm while assessing for pre-analytic factors which may affect its performance. Methods The decision support algorithm was used on digital pathology slides obtained from four different medical centers (A-D) and scanned by three different scanner models (by Philips, Hamamatsu and 3DHISTECH). A total of 192 cases and 1782 slides were used in this study. RGB histograms were constructed to compare images from the various medical centers and scanner models and highlight the differences in color and contrast. Results The algorithm was able to correctly identify ganglion cells in 99.2% of cases, from all medical centers (All scanned by the Philips slide scanner) as well as 95.5% and 100% of the slides scanned by the 3DHISTECH and Hamamatsu brand slide scanners, respectively. The total error rate for center D was lower than the other medical centers (3.9% vs 7.1%, 10.8% and 6% for centers A-C, respectively), the vast majority of errors being false positives (3.45% vs 0.45% false negatives). The other medical centers showed a higher rate of false negatives in relation to false positives (6.81% vs 0.29%, 9.8% vs 1.2% and 5.37% vs 0.63% for centers A-C, respectively). The total error rates for the Philips, Hamamatsu and 3DHISTECH brand scanners were 3.9%, 3.2% and 9.8%, respectively. RGB histograms demonstrated significant differences in pixel value distribution between the four medical centers, as well as between the 3DHISTECH brand scanner when compared to the Philips and Hamamatsu brand scanners. Conclusions The results reported in this paper suggest that the algorithm-based decision support system has sufficient robustness to be applicable for clinical practice. In addition, the novel method used in its development – Hierarchial-Contexual Analysis (HCA) may be applicable to the development of algorithm-assisted tools in other diseases, for which available datasets are limited. Validation of any given algorithm-assisted support system should nonetheless include data from as many medical centers and scanner models as possible.

show abstract

Section: Discussionmentioning

confidence: 99%

Algorithm-assisted diagnosis of Hirschsprung’s disease – evaluation of robustness and comparative image analysis on data from various labs and slide scanners

Greenberg,

Samueli,

Farkash

et al. 2024

Diagn Pathol

View full text Add to dashboard Cite

show abstract

“…Beyond identifying diagnoses via clinical images and electronic health record notes, machine learning techniques are being applied in dermatopathology ( 66 , 67 ). Groups have developed models to classify basal cell carcinoma in digitized Mohs micrographic surgery histology slides to reduce the workload of manually examining these slides ( 68 ).…”

Section: Applications Of Ai In Dermatologymentioning

confidence: 99%

Principles, applications, and future of artificial intelligence in dermatology

Omiye,

Gui,

Daneshjou

et al. 2023

Front. Med.

View full text Add to dashboard Cite

This paper provides an overview of artificial-intelligence (AI), as applied to dermatology. We focus our discussion on methodology, AI applications for various skin diseases, limitations, and future opportunities. We review how the current image-based models are being implemented in dermatology across disease subsets, and highlight the challenges facing widespread adoption. Additionally, we discuss how the future of AI in dermatology might evolve and the emerging paradigm of large language, and multi-modal models to emphasize the importance of developing responsible, fair, and equitable models in dermatology.

show abstract

“…Clearly labeling these pictures takes a lot of time and money, and usually requires the help of expert doctors. Not having accurate labels could make AI not learn correctly, resulting in good performance on training pictures but not in real-life situations [10,13].…”

Section: Numerous Pictures That Have Beenmentioning

confidence: 99%

Innovating Skin Cancer Detection with Transfer Learning, Feature Fusion, Efficient Data Management, Data Augmentation and Interdisciplinary Expertise for Better Outcomes

Turanli,

CIM

2023

Preprint

View full text Add to dashboard Cite

There is a growing acceptance of the use ofArtificial Intelligence (AI) in the use of healthcare,including in dermatology.AI is better at recognizingimages and learning patterns and due to this, i hasgrown into a powerful tool that can sometimes be moreaccurate than people and this has been demonstratedin recent studies showcasing the potential of AI.According to the World Cancer Research FundInternational, the global incidence of skin cancer maybe lesser than it is because of the challenges in datacollection, different types of skin cancer, interpretationof the results and demographics.Early detection and diagnosis in the majority of casessaves life. This has been proven overtime with mostNon-communicable diseases NCDs and this is nodifferent with skin cancer. This is where theapplication of AI in the diagnosis comes to play.In this study, we discussed the application of AI inMelanoma diagnosis. The use of machine learningmethods like Reinforcement Learning (RL), SupportVector Machines (SVM), K-Nearesr Neighbour (KNN)and Deep learning models like Convolutional NeuralNetwork (CNN), transfer learning and a combinationof both models were explored and showed significantaccuracies in Melanoma diagnosis compared to thetraditional methods. We also looked at the economicimpact of AI in healthcare, its cost-effectiveness andpublic perception of using AI.This study further emphasises the growing importanceof AI in healthcare and its integration into Melanomadiagnosis. It also explored the importance of AI inassisting dermatologists in the early detection of skincancer thereby saving lives.AI has the impact to makeour medical decision making easier, faster and withbetter accuracy and precision thereby increasingaccess to care.

show abstract

Deep learning in computational dermatopathology of melanoma: A technical systematic literature review

Cited by 14 publications

References 138 publications

Algorithm-assisted diagnosis of Hirschsprung’s disease – evaluation of robustness and comparative image analysis on data from various labs and slide scanners

Algorithm-assisted diagnosis of Hirschsprung’s disease – evaluation of robustness and comparative image analysis on data from various labs and slide scanners

Principles, applications, and future of artificial intelligence in dermatology

Innovating Skin Cancer Detection with Transfer Learning, Feature Fusion, Efficient Data Management, Data Augmentation and Interdisciplinary Expertise for Better Outcomes

Contact Info

Product

Resources

About