| ||||||
Skin Cancer Detection MethodSkin cancer is "the abnormal growth of abnormal cells" in the outer skin layer. When you have malignant skin cancer, you will notice some of the following conditions in your skin lesions, moles, spots;
If some of these conditions occur in your skin lesions and conditions do not improve, it's safer to see a doctor or dermatologist for diagnosis and treatment before too late. Skin Cancer Detection and Machine Learning - Deep Learning MethodIn combination with the above conditions, skin lesion photo images can be used to detect skin cancer using artificial intelligence machine learning techniques, specifically deep learning convolutional neural networks. To succeed machine learning, the followings should be noted;
It is noted that neural network training can take months or years of computation on high performance computer. Just executing Python code with a bunch of images won't produce good models. Systematic training data creation and methodological model training is needed. You need powerful dynamic tools with which you can guide training process like driving a car with dashboard information and handles. For example, CMSR Machine Learning Studio. Luck is very important part of neural network training as it relies on randomizations: random initialization of network weights and random shuffling of training data. Exceptionally good models can come with exceptional luck! Try multiple times to get that luck. About one million augmented images are needed for high accuracy on unseen images. But training such large data will be very time challenging. It can take years on average computers. Fast GPU with thousands of cores is recommended. Fast GPU can speed up training over hundreds or even thousands times fast. Regression vs ClassificationSkin cancer detection can be modeled as either regression or classification. Regression predicts probability of being skin cancer. Classification classifies as either true(=cancer) or false(=no cancer). It is noted that regression will have a single output variable while classification will have two output variables. Obviously training single output variable can be much easier than training two output variables. Regression will be more accurate and preferred. Dealing with Neural Network DivergenceNeural network training is a very tricky business. Neural network divergence is a common problem. There are two network divergence cases. One is Not a Number (NaN). This happens when absolute parameter values become too large. When this number is detected, network is dead. You will need to reinitialize network or abandon it. Another is that network accuracy drops suddenly and then unable to recover. Too high error correction ratio is the main reason of neural network divergence. By controlling error correction ratio, network divergence can be avoided. Network divergence is prominent with SGD. Mini batch size of 8 can avoid these problems while retaining SGD's traits. CMSR Machine Learning Studio keeps best versions. When divergence happens, you can return back to the best version which does not have divergence and keep training with lower error correction ratio. In this way, divergence can be overcome. The following figure shows that accuracy drops to the bottom. After returning to the best previous version with smaller error correction ratio, the network recovers. Significant level of augmentation also helps to avoid network divergence. Augmentation is essential. ![]() System AccuracyNeural network models used in this software were developed using very powerful machine learning modeling system CMSR Machine Learning Studio. Four deep convolutional neural network models, employing advanced proprietary algorithms and trained with 1 million photo images, are used. It has very high accuracy well over 90%. On 1 million training dataset images, 99.45% sensitivity and 97.52% specificity, 98.46% combined accuracy. (Note that sensitivity is the rate of detecting cancer as cancer. Specificity is the rate of detecting non-cancer as non-cancer.) On training dataset, 89.65% skin cancer images score as 99% risk probability. 96.90% of skin cancer images score as 90% or higher risk probability. Further tests on 1,853 skin cancer photos that were mostly not used in training showed over 97.4% detection rates. With good quality photos, detection rates will be much higher. Errors were mostly with poor or confusing photo quality. Notice that this is not 100% detection rate! So care should be taken. Accuracy of detecting not-cancer images as not-cancer is high. But less accurate than detecting cancer images. Note that some bad skin infection photos look alike cancer on visual level. So this system may rate as high risk. When conditions don't improve, always check with a doctor! LimitationsTesting revealed the following three main sources of errors;
The first problem is for users to improve photo quality. The other two problems have no clear solutions at the moment. In these cases, use this system with these in mind. [False negatives] Certain skin cancer looks similar to pimples or skin rash. In this case, this system may classify as low risk. Note that pimples and skin rash improve or disappear in days. If conditions don't improve in days, see a doctor or dermatologist. [False positives] Some bad skin infections look similar to skin cancer. This system may make mistakes. If conditions don't improve, you may need to see a doctor for treatment. Note that this system doesn't work well with dark skin. This is a limitation of this system. Skin Cancer Types DetectedAll three main types of skin cancer are detected: melanoma, basal cell carcinoma (BCC), and squamous cell carcinoma (SCC);
Merkel cell carcinoma (MCC) and Bowen's disease (BD) are also detected. Actinic keratosis (AK), which can become cancerous, is also detected. Photo RequirementsTo be accurate, well-focused close-up or well-focused high resolution color photos are recommended. Please take photos with these requirements in mind;
System RequirementsThis system runs on any JavaScript-enabled web browser. Your computer or mobile device may need at least 2GB RAM to enable JavaScript to process images. In addition, a digital camera or a camera enabled mobile device may be needed to take photos. About Data and PrivacyThis site does not collect data nor use cookies. Only the selected parts of images are sent to a server for analysis. Then discarded! Software VersionVersion 3.2 (3, January, 2023) | ||||||