A Dual Convolutional Neural Network Pipeline for Melanoma Diagnostics and Prognostics

Marie Bø-Sande, Edvin Benjaminsen, Neel Kanwal, Saul Fuster, Helga Hardardottir, Ingrid Lundal, Emilius A.M. Janssen, Kjersti Engan
Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL}), PMLR 233:20-26, 2024.

Abstract

Melanoma is a type of cancer that begins in the cells controlling the pigment of the skin, and it is often referred to as the most dangerous skin cancer. Diagnosing melanoma can be time-consuming, and a recent increase in melanoma incidents indicates a growing demand for a more efficient diagnostic process. This paper presents a pipeline for melanoma diagnostics, leveraging two convolutional neural networks, a diagnosis, and a prognosis model. The diagnostic model is responsible for localizing malignant patches across whole slide images and delivering a patient-level diagnosis as malignant or benign. Further, the prognosis model utilizes the diagnostic model’s output to provide a patient-level prognosis as good or bad. The full pipeline has an F1 score of 0.79 when tested on data from the same distribution as it was trained on.

Cite this Paper

BibTeX
@InProceedings{pmlr-v233-bo-sande24a, title = {A Dual Convolutional Neural Network Pipeline for Melanoma Diagnostics and Prognostics}, author = {B{\o}-Sande, Marie and Benjaminsen, Edvin and Kanwal, Neel and Fuster, Saul and Hardardottir, Helga and Lundal, Ingrid and Janssen, Emilius A.M. and Engan, Kjersti}, booktitle = {Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL})}, pages = {20--26}, year = {2024}, editor = {Lutchyn, Tetiana and Ramírez Rivera, Adín and Ricaud, Benjamin}, volume = {233}, series = {Proceedings of Machine Learning Research}, month = {09--11 Jan}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v233/bo-sande24a/bo-sande24a.pdf}, url = {https://proceedings.mlr.press/v233/bo-sande24a.html}, abstract = {Melanoma is a type of cancer that begins in the cells controlling the pigment of the skin, and it is often referred to as the most dangerous skin cancer. Diagnosing melanoma can be time-consuming, and a recent increase in melanoma incidents indicates a growing demand for a more efficient diagnostic process. This paper presents a pipeline for melanoma diagnostics, leveraging two convolutional neural networks, a diagnosis, and a prognosis model. The diagnostic model is responsible for localizing malignant patches across whole slide images and delivering a patient-level diagnosis as malignant or benign. Further, the prognosis model utilizes the diagnostic model’s output to provide a patient-level prognosis as good or bad. The full pipeline has an F1 score of 0.79 when tested on data from the same distribution as it was trained on.} }
Endnote
%0 Conference Paper %T A Dual Convolutional Neural Network Pipeline for Melanoma Diagnostics and Prognostics %A Marie Bø-Sande %A Edvin Benjaminsen %A Neel Kanwal %A Saul Fuster %A Helga Hardardottir %A Ingrid Lundal %A Emilius A.M. Janssen %A Kjersti Engan %B Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL}) %C Proceedings of Machine Learning Research %D 2024 %E Tetiana Lutchyn %E Adín Ramírez Rivera %E Benjamin Ricaud %F pmlr-v233-bo-sande24a %I PMLR %P 20--26 %U https://proceedings.mlr.press/v233/bo-sande24a.html %V 233 %X Melanoma is a type of cancer that begins in the cells controlling the pigment of the skin, and it is often referred to as the most dangerous skin cancer. Diagnosing melanoma can be time-consuming, and a recent increase in melanoma incidents indicates a growing demand for a more efficient diagnostic process. This paper presents a pipeline for melanoma diagnostics, leveraging two convolutional neural networks, a diagnosis, and a prognosis model. The diagnostic model is responsible for localizing malignant patches across whole slide images and delivering a patient-level diagnosis as malignant or benign. Further, the prognosis model utilizes the diagnostic model’s output to provide a patient-level prognosis as good or bad. The full pipeline has an F1 score of 0.79 when tested on data from the same distribution as it was trained on.
APA
Bø-Sande, M., Benjaminsen, E., Kanwal, N., Fuster, S., Hardardottir, H., Lundal, I., Janssen, E.A. & Engan, K.. (2024). A Dual Convolutional Neural Network Pipeline for Melanoma Diagnostics and Prognostics. Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL}), in Proceedings of Machine Learning Research 233:20-26 Available from https://proceedings.mlr.press/v233/bo-sande24a.html.

Related Material