Automated classification of urinary stones based on microcomputed tomography images using convolutional neural network

Fitri, L A; Haryanto, Freddy; Arimura, Hidetaka; Cui, Yunhao; Ninomiya, Kenta; Nakano, Risa; Haekal, Mohammad; Warty, Yuni; Fauzi, Umar

doi:10.1016/j.ejmp.2020.09.007

Cited by 27 publications

(11 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…All these preliminary and promising results described above explain why the medical community in urology is convinced of the interest of kidney stone recognition methods based on artificial intelligence Fitri et al (2020) and of the importance of incorporating computer aided diagnosis tools in their workflow Jahrreiss et al (2020). This work is an extension of a preliminary study Martínez et al (2020) which still improved the classification results using a RF classifier.…”

Section: Previous Attempts Of Kidney Stone Classificationmentioning

confidence: 98%

On the in vivo recognition of kidney stones using machine learning

Ochoa-Ruiz¹,

Estrade²,

López³

et al. 2022

Preprint

View full text Add to dashboard Cite

Determining the type of kidney stones allows urologists to prescribe a treatment to avoid recurrence of renal lithiasis. An automated in-vivo image-based classification method would be an important step towards an immediate identification of the kidney stone type required as a first phase of the diagnosis. In the literature it was shown on ex-vivo data (i.e., in very controlled scene and image acquisition conditions) that an automated kidney stone classification is indeed feasible. This pilot study compares the kidney stone recognition performances of six shallow machine learning methods and three deep-learning architectures which were tested with in-vivo images of the four most frequent urinary calculi types acquired with an endoscope during standard ureteroscopies. This contribution details the database construction and the design of the tested kidney stones classifiers. Even if the best results were obtained by the Inception v3 architecture (weighted precision, recall and F1-score of 0.97, 0.98 and 0.97, respectively), it is also shown that choosing an appropriate colour space and texture features allows a shallow machine learning method to approach closely the performances of the most promising deep-learning methods (the XGBoost classifier led to weighted precision, recall and F1-score values of 0.96). This paper is the first one that explores the most discriminant features to be extracted from images acquired during ureteroscopies.

show abstract

Section: Previous Attempts Of Kidney Stone Classificationmentioning

confidence: 98%

On the in vivo recognition of kidney stones using machine learning

Ochoa-Ruiz¹,

Estrade²,

López³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Ряд исследователей сообщает об определении состава камней непосредственно по данным КТ исследований в ручном режиме [13,14]. Од-нако, по мнению Fitri и соавт., подобная интерпретация является трудоемкой, она ассоциирована с наличием потенциальных погрешностей и всецело зависит от опыта и компетенций специалиста, что обуславливает актуальность разработки автоматизированных алгоритмов детекции, сегментации и определении химического состава камней [15].…”

Section: диджитализация в лечебно-диагностическом процессеunclassified

Artificial intelligence in the diagnosis and treatment of kidney stone disease.

Пранович¹,

Ismailov²,

Karelskaya³

et al. 2022

RJTEH

View full text Add to dashboard Cite

Complex health-care data will be growing exponentially as we enter the era of “big data”. Interpretation and analysis of a large amount of medical data for rapid and personalized decision-making require artificial intelligence-driven technologies and data mining techniques. Over the past decade, a rapid transition to the analysis, treatment and monitoring of kidney stone disease using the artificial intelligence (AI) with machine learning algorithms and artificial neural networks ensures the development of precise support systems. Machine learning algorithms able to predict the location of stones using computed tomography scans or ultrasound images, determine their chemical composition, recurrence rate based on demographic and genetic variables, spontaneous ureteral calculous passage, and treatment outcomes. Recent studies use rather simple machine learning methods and small training sets that reduce their accuracy, sensitivity, and specificity and limit their routine application. Therefore, the development of complex architectures, using multilayer artificial neural networks and convolutional neural networks, and integration other heterogeneous variables will improve the resultant predictive accuracy ensuring their introduction in real clinical practice.

show abstract

“…Spektrometer FTIR telah digunakan secara luas dalam identifikasi senyawa organik dan anorganik (Artz, et al, 2008) (Lopes, et al, 2018). Sedangkan identifikasi menggunakan CT-Scan memiliki kelebihan diantaranya memiliki resolusi spasial yang tinggi (Fitri, et al, 2020), dan mampu membedakan beberapa jenis batu kemih seperti calcium oxalate dengan phosphate (Duan, et al, 2015) Metodologi Sampel batu kemih yang digunakan dalam penelitian ini diperoleh dari rumah sakit Hasan Sadikin Bandung. Batu kemih tersebut memiliki ukuran dengan rentang 0,7 cm -3,2 cm dan tingkat kekerasan yang berbeda-beda.…”

Section: E-issn: 2355-8229unclassified

Application of uCT-scan and infrared spectroscopy for determination of urinary stone components

et al. 2021

Self Cite

View full text Add to dashboard Cite

Abstrak. Batu kemih merupakan salah satu penyakit dengan tingkat prevalensi yang cukup tinggi di Indonesia. Pengetahuan komposisi pada kandungan batu kemih dapat membantu tenaga medis dalam melakukan justifikasi penanganan tindakan lanjut pada pasien dengan tepat.Tujuan penelitian iniadalah menentukan kandungan mineral yang terdapat pada batu kemih menggunakan metode analisa spektrum inframerah dan nilai hounsfield unit (HU) yang terdapat pada citra yang dihasilkan dari modality mCT-Scan. Hasil karakterisasi fourier transform infrared spectroscopy (FTIR) memperlihatkan kandungan mineral batu kemih terdiri dari batu kemih calcium oxlate monohydrate, uric acid, batu campuran calcium oxalate dengan phosphate dan batu campuran cystine dengan phosphate sedangkan hasil dari scanning mCT memperlihatkan adanya kandungan mineral batu kemih campuran seperti batu campuran calcium oxalate dan cystine, batu campuran calcium oxalate, struvite, dan cystine, dan batu campuran calciumoxalate dan uric acid.Dari hasil penelitian ini dapat disimpulkan bahwa kedua modaliti tersebut mampu memperlihatkan kandungan mineral batu kemih dengan baik. Hal ini terlihat adanya spektrum serapan karakteristik dari FTIR setiap sampel berbeda-beda dan dari hasil citra mCT-Scan memperlihatkan nilai HU yang bervariasi sehingga mengindikasikan kandungan mineral pada sampel batu kemih yang diamati juga memiliki jenis yang berbeda-beda. Abstract. Urinary stones are a disease with a high prevalence rate in Indonesia. Knowledge of the composition of the urinary stone is an essential part to determine suitable treatments for patients. The aim of this research is to determine the mineral contained in urinary stones using the infrared spectrum and the value of HU (hounsfield unit) from the image mCT-Scan. The results of FTIR characterization showed that the mineral content of urinary stones consisted of calcium oxlate monohydrate, uric acid, calcium oxalate and phosphate mixed stones and cystine-phosphate mixed stones. mCT-Scan results showed the mineral content of urinary stones such as calcium oxalate and cystine mixed stone, calcium oxalate, struvite, and cystine mixed stones, and calcium oxalate and uric acid mixed stones. This show that the two modalities are be able to determine the mineral content of urinary stones. It can be seen that the characteristic absorption spectrum of the FTIR for each sample is different and from the mCT-Scan image results, the HU value varies so that it indicates the mineral content of the observed urine stone sample are different.

show abstract

Automated classification of urinary stones based on microcomputed tomography images using convolutional neural network

Cited by 27 publications

References 25 publications

On the in vivo recognition of kidney stones using machine learning

On the in vivo recognition of kidney stones using machine learning

Artificial intelligence in the diagnosis and treatment of kidney stone disease.

Application of uCT-scan and infrared spectroscopy for determination of urinary stone components

Contact Info

Product

Resources

About