A Probabilistic Framework for Detection of Skin Cancer by Raman Spectra

Sigurdur Sigurdsson

AbstractThis Ph.D. thesis focuses on objective methods for diagnosing skin cancer from Raman spectra. A method for suppressing background noise and dimension reduction in Raman spectra is suggested. A robust Bayesian framework for training a neural network is proposed, including an overfit control and outlier framework. Finally a visualization scheme for extracting important features from the trained neural network classifier based on sensitivity analysis is defined.

The performance on two types of skin cancer showed that 97.9% of basal cell carcinoma were identified correctly and 85.5% of malignant melanoma. The neural network classifier visualization showed that frequency bands, previously identified by visual inspection of Raman spectra by medical experts, were considered important for classification. Moreover, frequency band not previously used for skin lesion classification were identified. These identified important features are shown to originate from molecular structure changes in lipids and proteins.

While the theme of this dissertation is skin cancer diagnosis from Raman spectra, the dimension reduction and the neural network classifier can be applied in general to other types of pattern recognition problems.

In Danish:

Denne Ph.d. afhandling fokuserer på a objektive metoder til diagnosering af hudkræft fra Raman spektra. En metode til dæmpning af hud fluorisence og dimensions reduktion af Raman spektra er foreslået. En robust Bayesiansk fremgangsmåde til træning af neural netværk er foreslået, som indeholder outlier kontrol og overfitting håndtering. Endelig, er der defineret en visualisering metode af vigtige features fra det trænede neural netværk, baseret på indgangs/ udgangs følsomheds analyse.

Diagnose resultater for det neurale netværk for to typer hudkræft viser, at 97.9% af basal cell carcinoma og 85.5% af malignant melanoma er korrekt klassificeret. Visualisering viser at frekvensbånd i Raman spektra, som hudlæger havde identificeret som vigtige, også blev identificeret som vigtige af det neurale netværk. Endvidere, finder det neurale netværk frekvens bånd som ikke før er brugt til diagnosering af hudkræft. Disse vigtige frekvensbånd stammer fra forskel i molekyle struktur i lipider og proteiner.

Selv om temaet for denne afhandling er hudkræft diagnosering fra Raman spektra, kan dimensions reduceringen og det neurale netværk bruges til andre mønster genkendelses problemer uden videre tiltag.
TypePh.D. thesis [Academic thesis]
Year2003    pp. 202
PublisherInformatics and Mathematical Modelling, Technical University of Denmark, DTU
AddressRichard Petersens Plads, Building 321, DK-2800 Kgs. Lyngby
SeriesIMM-PHD-2003-114
NoteSupervisor: Lars Kai Hansen
Electronic version(s)[zip]
BibTeX data [bibtex]
IMM Group(s)Intelligent Signal Processing


Back  ::  IMM Publications