FreshRSS

🔒
❌ Acerca de FreshRSS
Hay nuevos artículos disponibles. Pincha para refrescar la página.
AnteayerTus fuentes RSS

Deep learning model to predict lupus nephritis renal flare based on dynamic multivariable time-series data

Por: Huang · S. · Chen · Y. · Song · Y. · Wu · K. · Chen · T. · Zhang · Y. · Jia · W. · Zhang · H.-T. · Liang · D.-D. · Yang · J. · Zeng · C.-H. · Li · X. · Liu · Z.-H.
Objectives

To develop an interpretable deep learning model of lupus nephritis (LN) relapse prediction based on dynamic multivariable time-series data.

Design

A single-centre, retrospective cohort study in China.

Setting

A Chinese central tertiary hospital.

Participants

The cohort study consisted of 1694 LN patients who had been registered in the Nanjing Glomerulonephritis Registry at the National Clinical Research Center of Kidney Diseases, Jinling Hospital from January 1985 to December 2010.

Methods

We developed a deep learning algorithm to predict LN relapse that consists of 59 features, including demographic, clinical, immunological, pathological and therapeutic characteristics that were collected for baseline analysis. A total of 32 227 data points were collected by the sliding window method and randomly divided into training (80%), validation (10%) and testing sets (10%). We developed a deep learning algorithm-based interpretable multivariable long short-term memory model for LN relapse risk prediction considering censored time-series data based on a cohort of 1694 LN patients. A mixture attention mechanism was deployed to capture variable interactions at different time points for estimating the temporal importance of the variables. Model performance was assessed according to C-index (concordance index).

Results

The median follow-up time since remission was 4.1 (IQR, 1.7–6.7) years. The interpretable deep learning model based on dynamic multivariable time-series data achieved the best performance, with a C-index of 0.897, among models using only variables at the point of remission or time-variant variables. The importance of urinary protein, serum albumin and serum C3 showed time dependency in the model, that is, their contributions to the risk prediction increased over time.

Conclusions

Deep learning algorithms can effectively learn through time-series data to develop a predictive model for LN relapse. The model provides accurate predictions of LN relapse for different renal disease stages, which could be used in clinical practice to guide physicians on the management of LN patients.

Epidemiology, site-specific characteristics and survival of carcinosarcoma: a retrospective study based on SEER database

Por: Tang · L.-S. · Zhou · Y.-W. · Wang · J.-L. · Zhang · G.-X. · Xu · C.-H. · Liu · J.-Y. · Qiu · M.
Objectives

Carcinosarcoma (CS) is a rare and biphasic malignancy characterised by a highly invasive biological nature and poor prognosis. This study explored the epidemiology, site-specific characteristics and survival outcome of CS.

Design

We conducted a retrospective study in the Surveillance, Epidemiology and End Results (SEER) database (1975–2018) for primary CS.

Setting and participants

SEER database includes publicly available information from regional and state cancer registries in the US centres. A total of 5042 CS patients were identified. We selected the top five anatomic CS (uterus, double adnexa, lung, bladder and breast) patients for further analysis.

Primary outcome measures

Incidence was estimated by geographical region, age, sex, race, stage and primary site. Trends were calculated using joinpoint regression. The cancer-specific survival (CSS) rate and initial treatment were summarised.

Results

Nearly 80% of CS occurred in the uterus and double adnexa, followed by lung, bladder and breast. The elderly and black population presented the highest age-adjusted rate of CS. The rates of distant metastasis in CS progressively increased from 1989 to 2018. Atlanta was the area with the highest incidence at 0.7 per 100 000. Pulmonary and bladder CS more frequently occurred in men and were diagnosed with regional stage. Distant metastasis was mostly found in ovary/fallopian tube CS. Radiotherapy was more commonly applied in uterine CS, while adnexa CS cases were more likely to receive chemotherapy. Multiple treatments were more used in breast CS. Pulmonary CS seemed to suffer worse CSS (median: 9.92 months), for which radiotherapy might not provide survival benefits (HR 0.60, 95% CI 0.42 to 0.86). Compared with the common histological types in each site, CS had the shortest survival.

Conclusions

CS has unique clinical features in each primary site. Substantial prognosis variances exist based on tumour locations. The aggressive course is the common feature in CS at all sites.

❌