Abstract
PurposeRecent advances in machine learning have enabled better understanding of large and complex visual data. Here, we aim to investigate patient outcome prediction with a machine learning method using only an image of tumour sample as an input.MethodsUtilising tissue microarray (TMA) samples obtained from the primary tumour of patients (N = 1299) within a nationwide breast cancer series with long-term-follow-up, we train and validate a machine learning method for patient outcome prediction. The prediction is performed by classifying samples into low or high digital risk score (DRS) groups. The outcome classifier is trained using sample images of 868 patients and evaluated and compared with human expert classification in a test set of 431 patients.ResultsIn univariate survival analysis, the DRS classification resulted in a hazard ratio of 2.10 (95% CI 1.33–3.32, p = 0.001) for breast cancer-specific survival. The DRS classification remained as an independent predictor of breast cancer-specific survival in a multivariate Cox model with a hazard ratio of 2.04 (95% CI 1.20–3.44, p = 0.007). The accuracy (C-index) of the DRS grouping was 0.60 (95% CI 0.55–0.65), as compared to 0.58 (95% CI 0.53–0.63) for human expert predictions based on the same TMA samples.ConclusionsOur findings demonstrate the feasibility of learning prognostic signals in tumour tissue images without domain knowledge. Although further validation is needed, our study suggests that machine learning algorithms can extract prognostically relevant information from tumour histology complementing the currently used prognostic factors in breast cancer.
Highlights
There is a growing interest around the potential of machine learning to improve the accuracy of medical diagnostics [1]
The digital risk score (DRS) model performance rates measured with area under receiver operating characteristics curve (AUC) on the test and training sets were 0.58 and 0.63, respectively, indicating no substantial model overfitting (Supplementary Fig. S1)
We found that by utilising machine learning algorithms it is possible to extract information relevant for breast cancer patient outcomes from tumour tissue images stained for the basic morphology only
Summary
There is a growing interest around the potential of machine learning to improve the accuracy of medical diagnostics [1]. Novel machine learning techniques have advanced the state-of-the-art in several pattern recognition tasks [2, 3], and have the potential to extract clinically relevant information from complex medical imaging data sets. Methods using deep learning have been successful in various medical image analysis tasks [4, 5], some of them reaching performance of experienced specialists in individual diagnostic tasks [6, 7]. Whole-slide scanners have enabled accurate digitisation of histological samples with sub-micrometre resolution allowing for computerised analysis of the specimens with machine learning algorithms [8]. Recent reviews [5, 8, 19] offer thorough summaries on methods developed for analysis of histological samples
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
