Investigation of the effect of serum added to the medium on the classification success in the detection of melanoma using surface-enhanced Raman spectroscopy

Fatma Uysal Çiloğlu

Article Details

Investigation of the effect of serum added to the medium on the classification success in the detection of melanoma using surface-enhanced Raman spectroscopy

Authors:

Fatma Uysal Çiloğlu

Publisher:

Nigde Omer Halisdemir University

Main Topics:

Makine Öğrenme (Diğer) | Biyomedikal Bilimler ve Teknolojiler | Biyomedikal Mühendisliği (Diğer) | Nanofotonik

Journal:

Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 1, pp.70-79

Document Type:

Article

Publication Date:

1/2025

Document Language:

Turkish

Document ID (Org No.):

2001339028

pISSN:

2564-6605

Keywords:

Abstract:

 Due to the disadvantages of the methods used for the detection of melanoma, which ranks among the leading causes of death worldwide, there is a need for the development of new methods. In this context, surface-enhanced Raman spectroscopy SERS emerges as a powerful and alternative technique. In this study, the effect of serum added to the cell culture medium during the culture of skin fibroblasts, tumor-associated fibroblasts, and melanoma cells on the classification success of SERS spectra collected from the cells was investigated. Feature extraction was performed using principal component analysis PCA, reconstruction-independent component analysis RICA, and sparse filtering SF methods on SERS spectra obtained from cells cultured in media with and without serum. Support vector machines SVM and k-nearest neighbors KNN algorithms were used to classify the extracted features. It was found that the classification accuracy of spectra collected from cells cultured in serum-free media was higher compared to those from the serum-added group when classified using machine learning techniques. The classification accuracy of the serum-free cell group was found to be 96.40.4 using SF with SVM and 96.80.29 using PCA with the KNN algorithm, with a statistically significant difference in classification accuracy between the serum-added and serum-free groups.

 Due to the disadvantages of the methods used for the detection of melanoma, which ranks among the leading causes of death worldwide, there is a need for the development of new methods. In this context, surface-enhanced Raman spectroscopy SERS emerges as a powerful and alternative technique. In this study, the effect of serum added to the cell culture medium during the culture of skin fibroblasts, tumor-associated fibroblasts, and melanoma cells on the classification success of SERS spectra collected from the cells was investigated. Feature extraction was performed using principal component analysis PCA, reconstruction-independent component analysis RICA, and sparse filtering SF methods on SERS spectra obtained from cells cultured in media with and without serum. Support vector machines SVM and k-nearest neighbors KNN algorithms were used to classify the extracted features. It was found that the classification accuracy of spectra collected from cells cultured in serum-free media was higher compared to those from the serum-added group when classified using machine learning techniques. The classification accuracy of the serum-free cell group was found to be 96.40.4 using SF with SVM and 96.80.29 using PCA with the KNN algorithm, with a statistically significant difference in classification accuracy between the serum-added and serum-free groups.

 Due to the disadvantages of the methods used for the detection of melanoma, which ranks among the leading causes of death worldwide, there is a need for the development of new methods. In this context, surface-enhanced Raman spectroscopy SERS emerges as a powerful and alternative technique. In this study, the effect of serum added to the cell culture medium during the culture of skin fibroblasts, tumor-associated fibroblasts, and melanoma cells on the classification success of SERS spectra collected from the cells was investigated. Feature extraction was performed using principal component analysis PCA, reconstruction-independent component analysis RICA, and sparse filtering SF methods on SERS spectra obtained from cells cultured in media with and without serum. Support vector machines SVM and k-nearest neighbors KNN algorithms were used to classify the extracted features. It was found that the classification accuracy of spectra collected from cells cultured in serum-free media was higher compared to those from the serum-added group when classified using machine learning techniques. The classification accuracy of the serum-free cell group was found to be 96.40.4 using SF with SVM and 96.80.29 using PCA with the KNN algorithm, with a statistically significant difference in classification accuracy between the serum-added and serum-free groups.

 Dünya genelinde ölüm oranları açısından ilk sıralarda yer alan melanomanın tespitinde kullanılan yöntemlerin dezavantajlarından dolayı yeni yöntemlerin geliştirilmesine ihtiyaç duyulmaktadır. Bu doğrultuda yüzeyde zenginleştirilmiş Raman spektroskopisi YZRS güçlü ve alternatif bir teknik olarak karşımıza çıkmaktadır. Bu çalışmada deri fibroblast, tümör ilişkili fibroblast ve melanoma hücrelerinin kültür aşamasında hücre kültür ortamına eklenen serumun hücrelerden toplanan YZRS spektrumlarının sınıflandırma başarısı üzerine etkisi araştırılmıştır. Serum eklenen ve eklenmeyen ortamda kültürlenen hücrelerden elde edilen YZRS spektrumlarının; temel bileşen analizi TBA, yeniden yapılanma bağımsız bileşen analizi YYBBA ve seyrek filtreleme SF yöntemleri kullanılarak öz nitelik çıkarımı gerçekleştirilmiştir. Çıkarılan öz nitelikleri sınıflandırmak için destek vektör makinası DVM ve k en yakın komşu KEYK algoritması kullanılmıştır. Serum eklenmeyen hücre kültür ortamlarında kültürlenen hücrelerden toplanan spektrumların makine öğrenme teknikleri ile sınıflandırma işlemi gerçekleştirildiğinde sınıflandırma başarılarının serum eklenen gruba göre daha yüksek olduğu tespit edilmiştir. Serum eklenmeyen hücre grubunun sınıflandırma başarısı SF ile DVM kullanıldığında 96.40.4, TBA ile KEYK algoritması kullanıldığında 96.80.29 olarak bulunmuş olup serum eklenen ve eklenmeyen gruplara ait sınıflandırma başarısı arasında istatistiksel açıdan anlamlı bir fark bulunmuştur.

Article Details

Investigation of the effect of serum added to the medium on the classification success in the detection of melanoma using surface-enhanced Raman spectroscopy

Authors:

Fatma Uysal Çiloğlu

Publisher:

Nigde Omer Halisdemir University

Main Topics:

Makine Öğrenme (Diğer) | Biyomedikal Bilimler ve Teknolojiler | Biyomedikal Mühendisliği (Diğer) | Nanofotonik

Journal:

Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 1, pp.70-79

Document Type:

Article

Publication Date:

1/2025

Document Language:

Turkish

Document ID (Org No.):

2001339028

pISSN:

2564-6605

melanoma

Surface enhanced Raman spectroscopy

Principal component analysis

Reconstruction independent component analysis

sparse filteringmelanoma

Yüzeyde zenginleştirilmiş Raman spektroskopi

temel bileşen analizi

yeniden yapılandırma bağımsız bileşen analizi

seyrek filtreleme

Issues

Article Details

Authors: Fatma Uysal Çiloğlu

Publisher: Nigde Omer Halisdemir University

Main Topics: Makine Öğrenme (Diğer) | Biyomedikal Bilimler ve Teknolojiler | Biyomedikal Mühendisliği (Diğer) | Nanofotonik

Journal: Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 1, pp.70-79

Document Type: Article

Publication Date: 1/2025

Document Language: Turkish

Document ID (Org No.): 2001339028

pISSN: 2564-6605

Issues

Authors:

Fatma Uysal Çiloğlu

Publisher:

Nigde Omer Halisdemir University

Main Topics:

Makine Öğrenme (Diğer) | Biyomedikal Bilimler ve Teknolojiler | Biyomedikal Mühendisliği (Diğer) | Nanofotonik

Journal:

Niğde Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi Vol.14 Issue 1, pp.70-79

Document Type:

Article

Publication Date:

1/2025

Document Language:

Turkish

Document ID (Org No.):

2001339028

pISSN:

2564-6605