Automatic Detection of Cervical Region from VIA and VILI Images using Machine Learning

Kanitkar, Ashutosh; Kulkarni, Rutwik; Joshi, Varun; Karwa, Yash; Gindi, Sanjyot; Kale, Geetanjali

doi:10.1109/cse/euc.2019.00010

Cited by 2 publications

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The potential for automatic analysis of digital cervical images in revolutionizing screening for precancers has motivated the development of several automatic and semiautomatic image analysis algorithms. These include algorithms for anatomical landmark detection [9], cervix region detection [10], [11], cervix type detection [12], pre-cancerous lesion detection-segmentation [13]- [15] and disease diagnosis [16]- [18]. Since our main concern is detecting precancer (or worst disease condition) in cervical images, we restrict our literature review to the topically relevant algorithms.…”

Section: Related Literaturementioning

confidence: 99%

Deep Metric Learning for Cervical Image Classification

et al. 2021

View full text Add to dashboard Cite

Cervical cancer is caused by the persistent infection of certain types of the Human Papillomavirus (HPV) and is a leading cause of female mortality particularly in low and middle-income countries (LMIC). Visual inspection of the cervix with acetic acid (VIA) is a commonly used technique in cervical screening. While this technique is inexpensive, clinical assessment is highly subjective, and relatively poor reproducibility has been reported. A deep learning-based algorithm for automatic visual evaluation (AVE) of aceto-whitened cervical images was shown to be effective in detecting confirmed precancer (i.e. direct precursor to invasive cervical cancer). The images were selected from a large longitudinal study conducted by the National Cancer Institute in the Guanacaste province of Costa Rica.The training of AVE used annotation for cervix boundary, and the data scarcity challenge was dealt with manually optimized data augmentation. In contrast, we present a novel approach for cervical precancer detection using a deep metric learning-based (DML) framework which does not incorporate any effort for cervix boundary marking. The DML is an advanced learning strategy that can deal with data scarcity and bias training due to class imbalance data in a better way. Three different widely-used state-of-the-art DML techniques are evaluated-(a) Contrastive loss minimization, (b) N-pair embedding loss minimization, and, (c) Batch-hard loss minimization. Three popular Deep Convolutional Neural Networks (ResNet-50, MobileNet, NasNet) are configured for training with DML to produce class-separated (i.e. linearly separable) image feature descriptors. Finally, a K-Nearest Neighbor (KNN) classifier is trained with the extracted deep features. Both the feature quality and classification performance are quantitatively evaluated on the same data set as used in AVE. It shows that, unlike AVE, without using any data augmentation, the best model produced from our research improves specificity in disease detection without compromising sensitivity. The present research thus paves the way for new research directions for the related field.

show abstract

Section: Related Literaturementioning

confidence: 99%

Deep Metric Learning for Cervical Image Classification

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Untuk mengatasi hal tersebut berbagai penelitian telah dilakukan seperti pada penelitian [7] dimana Acetowhite Epithelium Zone (AEZ) yang merupakan lesi IVA positif (lesi kanker) dapat dideteksi secara semi-otomatis dengan menggunakan metode registered ratio image. Beberapa penelitian deteksi kanker serviks dengan pengolahan citra digital dilakukan dengan artificial intelligence baik dengan platform machine learning [8]- [10] maupun dengan deep learning [11], [12]. Untuk meningkatkan performa system deteksi kanker serviks dengan pengolahan citra digital telah dilakukan dengan meningkatkan kualitas citra input [13] dan optimasi algoritma klasifikasi system [14], [15].…”

unclassified

Sistem Deteksi Pra-Kanker Serviks dengan Pengolahan Citra Hasil Inspeksi Visual Asam Asetat

Fauzi

Surya

Magdalena

et al. 2021

View full text Add to dashboard Cite

Kanker serviks merupakan penyakit mematikan nomor satu di Indonesia dengan angka kematian tertinggi pada wanita. Berbagai upaya untuk mengurangi angka kematian wanita Indonesia akibat kanker serviks telah banyak dilakukan, salah satunya dengan melakukan screening kanker menggunakan tes inspeksi visual asam asetat (tes IVA). Tes ini merupakan upaya screening untuk mengetahui pra-cancer atau invasive cancer pada kanker serviks dengan memunculkan Acetowhite Epithelium Zone (AEZ) yang dapat dikategorikan sebagai lesi IVA positif maupun lesi jinak. Umumnya, AEZ dapat dilihat dengan kasat mata yang memerlukan keahlian khusus sehingga hasil pengamatannya akan bersifat subjektif dan bergantung pada pengalaman operator. Selain itu, utilitas pemeriksaan kanker serviks ini pun dinilai terbatas dikarenakan sedikitnya jumlah operator ahli yang terlatih. Pada penelitian ini, lesi pra-kanker serviks dikuantifikasi dengan pengolahan citra digital. Citra yang digunakan adalah citra hasil inspeksi visual asam asetat atau citra area mulut rahim yang telah diolesi oleh asam asetat dan dinyatakan terdapat sambungan skuamosa kolumnar (SSK) positif. Kuantifikasi citra lesi pra-kanker serviks dilakukan dengan menggunakan metode standarisasi karakter warna citra pada RGB dan HSV. Pengujian system deteksi lesi pra-kanker serviks diukur dengan menggunakan parameter akurasi, sensitivitas dan spesifisitas terhadap pengaruh tingkat kecerahan dan mean filter. Melalui penelitian ini didapatkan klasifikasi citra tes IVA beserta area lesi IVA positif yang optimal dengan tingkat akurasi 81%, nilai sensitivitas 78% dan nilai spesifisitas 84%. Performa system sangat dipengaruhi oleh ketajaman dan efek pencahayaan pada citra, baik itu intensitas cahaya, efek bayangan, maupun efek pantulan cahaya.

show abstract

Automatic Detection of Cervical Region from VIA and VILI Images using Machine Learning

Cited by 2 publications

References 9 publications

Deep Metric Learning for Cervical Image Classification

Deep Metric Learning for Cervical Image Classification

Sistem Deteksi Pra-Kanker Serviks dengan Pengolahan Citra Hasil Inspeksi Visual Asam Asetat

Contact Info

Product

Resources

About