Feasibility study of hyperspectral colposcopy as a novel tool for detecting precancerous cervical lesions

Sep 23, 2026

Uploaded by Jiven

Feasibility study proposes hyperspectral colposcopy to improve detection of precancerous cervical lesions by combining imaging and spectroscopy. A custom hyperspectral colposcope collected data from 62 patients over six months, followed by PCA-based preprocessing for unsupervised cervical-region segmentation. The workflow extracted spectral signatures from different tissue types, and statistical analysis identified measurable differences and tissue alterations. The findings indicate a promising path toward higher-precision cervical lesion diagnosis beyond conventional screening and colposcopy limitations.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages15

Reading Time38 min

AccessFree to download

Introduction

Cervical cancer burden and screening limitations

Natural history and CIN classification

Pap smear and colposcopy workflow

Motivation for objective imaging tools and deep learning

Hyperspectral imaging approach in this study

[www. nature.com/scientificreports](www. nature.com/scientificreports) OPEN Feasibility study of hyperspectral colposcopy as a novel tool for detecting precancerous cervical lesions Carlos Vega1􀀍, Norberto Medina2, Laura Quintana-Quintana1, Raquel Leon1, Himar Fabelo1,3,4, Jorge Rial5, Alicia Martín2 & Gustavo M. Callico1 Cervical cancer remains a major global health concern, with a specially alarming incidence in younger women. Traditional detection techniques such as the Pap smear and colposcopy often lack sensitivity and specificity and are highly dependent on the experience of the gynaecologist. In response, this study proposes the use of Hyperspectral Imaging, a pioneering technology that combines traditional imaging with spectroscopy to provide detailed spatial and spectral information. Over a period of six-months, our custom-designed hyperspectral colposcope was used on 62 patients. The gathered data underwent a specialized preprocessing workflow using a PCA-based strategy for unsupervised segmentation of the cervical region. This process extracted spectral signatures from various tissue types, and our subsequent statistical analysis highlighted its ability to detect differences and alterations in the cervical tissue. This offers a promising avenue for improving the precision of cervical lesion diagnosis. Cervical cancer is recognized as a major global health concern, ranking fourth in terms of incidence and third in mortality, among women worldwide in 2020 and affecting younger populations under 45 years old at an alarming rate1. In Europe, the incidence of cervical cancer is 15 cases per 100,000 inhabitants, although there is a high variability between regions2. For example, in Canary Islands (Spain), where this study was conducted, the age-standardized incidence of cervical cancer in 2018 was 8.5 cases per 100,000 inhabitants, compared with a national average standardized incidence of 5.4 cases per 100,000 inhabitants3,4. These data highlight the necessity for enhanced detection and screening tools for cervical cancer, especially at early stages4. The invasive squamous epithelial carcinoma of the cervix is developed from precursor lesions, specifically the high-grade squamous Cervical Intraepithelial Neoplasia (CIN) . For this pathology to grow, the patient must have been persistently infected with the Human Papilloma Virus (HPV)5, although this is not a sufficient condition for its development. The natural history of cervical cancer starts several years or decades earlier in the patient, with the HPV infection in the keratinocytes ofthe basal stratum of the squamous epithelium. Initially, low-grade benign lesions appear, which progress to squamous cell high-grade lesions with a high potential for malignancy6. CIN can be classified on a scale from one to three, depending on how deeply the abnormal cells affect the squamous epithelium. Destruction and involvement of the lower third of the squamous epithelium is called CIN 1, a benign lesion that regresses spontaneously in 60% of cases6. In the cases where the degree of involvement of the squamous epithelium affects two-thirds, it is called CIN 2. Finally, if the abnormal cells affect more than two-thirds ofthe squamous epithelium, it is called CIN 3. Both, CIN 2 and CIN 3 lesions, are identified as high-grade lesions. The Papanicolaou test (Pap smear)7 is a primary cervical screening method to detect potential precancerous and cancerous lesions of the cervix and is used in the early detection of possible CIN. The test is performed by collecting cells from the cervix with a cytobrush and examining the samples under a microscope8. Cytology is usually combined with an HPV test to determine if the patient is infected with the virus and which type. If the test shows abnormal results, the patient undergoes a colposcopic examination. The colposcope is a microscope with stereoscopic vision, low magnification, and a powerful white light source. To facilitate diag

FAQs

What problem does the study address in cervical lesion detection?

Traditional approaches such as Pap smear and colposcopy can lack sensitivity and specificity and depend heavily on clinician experience. The study targets the need for more objective and precise detection tools.

How was the hyperspectral colposcope used in the study?

A custom-designed hyperspectral colposcope was used over six months on 62 patients to collect hyperspectral data from cervical regions.

What data processing method was applied to the collected hyperspectral data?

The study used a PCA-based strategy for unsupervised segmentation of the cervical region, extracting spectral signatures from different tissue types.

Sep 23, 2026

Uploaded by Jiven