Smartphone-based fluorescence spectroscopic device for cervical precancer diagnosis: a random forest classification of in vitro data

Shivam Shukla; Chaitanya Vishwakarma; Amar Nath Sah; Shikha Ahirwar; Kiran Pandey; Asima Pradhan; Asima Pradhan; Asima Pradhan

doi:10.1364/AO.496543

Applied Optics
Vol. 62,
Issue 25,
pp. 6826-6834
(2023)
•https://doi.org/10.1364/AO.496543

Smartphone-based fluorescence spectroscopic device for cervical precancer diagnosis: a random forest classification of in vitro data

Shivam Shukla, Chaitanya Vishwakarma, Amar Nath Sah, Shikha Ahirwar, Kiran Pandey, and Asima Pradhan

Not Accessible

Your library or personal account may give you access

Get PDF
Email
Share
Get Citation
Copy Citation Text
Shivam Shukla, Chaitanya Vishwakarma, Amar Nath Sah, Shikha Ahirwar, Kiran Pandey, and Asima Pradhan, "Smartphone-based fluorescence spectroscopic device for cervical precancer diagnosis: a random forest classification of in vitro data," Appl. Opt. 62, 6826-6834 (2023)

Export Citation
- BibTex
- Endnote (RIS)
- HTML
- Plain Text
Citation alert
Save article
Editors' Pick

Check for updates

Abstract

Cervical cancer can be treated and cured if diagnosed at an early stage. Optical devices, developed on smartphone-based platforms, are being tested for this purpose as they are cost-effective, robust, and field portable, showing good efficiency compared to the existing commercial devices. This study reports on the applicability of a 3D printed smartphone-based spectroscopic device (3D-SSD) for the early diagnosis of cervical cancer. The proposed device has the ability to evaluate intrinsic fluorescence (IF) from the collected polarized fluorescence (PF) and elastic-scattering (ES) spectra from cervical tissue samples of different grades. IF spectra of 30 cervical tissue samples have been analyzed and classified using a combination of principal component analysis (PCA) and random forest (RF)-based multi-class classification algorithm with an overall accuracy above 90%. The usage of smartphone for image collection, spectral data analysis, and display makes this device a potential contender for use in clinics as a regular screening tool.

Full Article | PDF Article

More Like This

Deep learning and sub-band fluorescence imaging-based method for caries and calculus diagnosis embeddable on different smartphones

Cheng Wang, Rongjun Zhang, Xiaoling Wei, Le Wang, Peiyu Wu, and Qi Yao
Biomed. Opt. Express 14(2) 866-882 (2023)

Comparison of whole blood and serum samples of breast cancer based on laser-induced breakdown spectroscopy with machine learning

Bushra Sana Idrees, Geer Teng, Ayesha Israr, Huma Zaib, Yasir Jamil, Muhammad Bilal, Sajid Bashir, M. Nouman Khan, and Qianqian Wang
Biomed. Opt. Express 14(6) 2492-2509 (2023)

Intelligent smartphone-based multimode imaging otoscope for the mobile diagnosis of otitis media

Thiago C. Cavalcanti, Hah Min Lew, Kyungsu Lee, Sang-Yeon Lee, Moo Kyun Park, and Jae Youn Hwang
Biomed. Opt. Express 12(12) 7765-7779 (2021)

Previous Article Next Article

Supplementary Material (1)

Name	Description
Supplement 1	Supplementary document

Data availability

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.

Cited By

You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Figures (7)

You do not have subscription access to this journal. Figure files are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Tables (7)

You do not have subscription access to this journal. Article tables are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Equations (5)

You do not have subscription access to this journal. Equations are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Sr.No.	Category	No. of Patients	No. of Sites
1	Normal	10	40
2	CINI	15	45
3	CINII	5	15

Principal Components	Eigen Value	Variance (%)	Cumulative Variance (%)
1	716.32	76.78	76.78
2	90.73	9.72	86.51
3	25.70	2.75	89.27
4	19.64	2.11	91.37
5	11.17	1.20	92.57
6	9.87	1.10	93.63
7	6.96	0.75	94.37
8	6.33	0.68	95.05
9	5.32	0.57	95.62
10	4.78	0.51	96.14

Sr. No.	Hyperparameter	Value/Method
1	max_depth	30
2	min_sample_split	15
1	min_samples_leaf	15
1	n_estimators	100
1	max_features	auto, sqrt, log2
1	bootstrap	true
3	Criterion	gini

Sr.		Training	Test Score
No.	Approach	Score (%)	(%)	RMSE
1	RF	100	88.89	0.4266
2	RF $+$ GridSearchCV	100	91.11	0.3651
3	RF $+$ RandomizedSearchCV	100	95.56	0.2108

		CINI Versus	CINII Versus	CINII Versus
Sr. No.	Parameter	Normal	CINI	Normal
1	Accuracy	93	92	96
2	Sensitivity	84	100	100
3	Specificity	100	84	93
4	PPA	100	88	94
5	NPA	88	100	100

Abstract

Supplementary Material (1)

Data availability

Cited By

Figures (7)

Tables (7)

Equations (5)

Applied Optics