arXiv subjects

Rana Jafari

Publications and source records attributed to Rana Jafari.

6 recordsLinked to original sources

arXiv2026

Quantitative Pulse Shape-Instability Analysis Using 2D-Runs FROG

We present a method for quantifying pulse-shape instability in a train of pulses using multi-shot Second-Harmonic-Generation Frequency-Resolved Optical Gating (SHG FROG). All versions of multi-shot FROG have previously shown the ability to distinguish stable from unstable pulse trains, as systematic differences appear between measured and retrieved traces when instability is present. This has proved possible because the recently introduced Retrieved-Amplitude N-grid Algorithmic (RANA) approach provides highly reliable pulse retrieval, even for unstable pulse trains and in the presence of noise, thus eliminating the possibility that algorithm stagnation, which mimics the effects of pulse-shape instability, could be confused for it. In other words, RANAs excellent performance ensures that any non-random discrepancies between measured and retrieved FROG traces reflect physical pulse-shape instability, rather than algorithmic stagnation. To begin to quantify such instability, we now introduce an instability parameter, R. It involves the use of the well-known statistical Runs test, which tests for systematic error in fits to one-dimensional (1D) data. A runs test counts the runs consecutive points in the plot of the difference between the data and fit with the same sign evaluating the goodness of the fit while minimizing the effects of random error. However, because FROG traces are functions of two variables, we must extend the usual 1D runs test to two dimensions, that is, to enumerate 2D runs hills and valleys in the difference between measured and retrieved 2D FROG traces. Many small 2D runs indicate only random noise-like differences and hence a stable pulse train, whereas few large runs reflect additional systematic error and hence pulse-shape instability.

physics.optics↗

arXiv2023

MDACE: MIMIC Documents Annotated with Code Evidence

We introduce a dataset for evidence/rationale extraction on an extreme multi-label classification task over long medical documents. One such task is Computer-Assisted Coding (CAC) which has improved significantly in recent years, thanks to advances in machine learning technologies. Yet simply predicting a set of final codes for a patient encounter is insufficient as CAC systems are required to provide supporting textual evidence to justify the billing codes. A model able to produce accurate and reliable supporting evidence for each code would be a tremendous benefit. However, a human annotated code evidence corpus is extremely difficult to create because it requires specialized knowledge. In this paper, we introduce MDACE, the first publicly available code evidence dataset, which is built on a subset of the MIMIC-III clinical records. The dataset -- annotated by professional medical coders -- consists of 302 Inpatient charts with 3,934 evidence spans and 52 Profee charts with 5,563 evidence spans. We implemented several evidence extraction methods based on the EffectiveCAN model (Liu et al., 2021) to establish baseline performance on this dataset. MDACE can be used to evaluate code evidence extraction methods for CAC systems, as well as the accuracy and interpretability of deep learning models for multi-label classification. We believe that the release of MDACE will greatly improve the understanding and application of deep learning technologies for medical coding and document classification.

cs.CL↗

arXiv2020

Linear encoding of the spatiotemporal cat

The dynamics of an extended, spatiotemporally chaotic system might appear extremely complex. Nevertheless, the local dynamics, observed through a finite spatiotemporal window, can often be thought of as a visitation sequence of a finite repertoire of finite patterns. To make statistical predictions about the system, one needs to know how often a given pattern occurs. Here we address this fundamental question within a spatiotemporal cat, a 1-dimensional spatial lattice of coupled cat maps evolving in time. In spatiotemporal cat, any spatiotemporal state is labeled by a unique 2-dimensional lattice of symbols from a finite alphabet, with the lattice states and their symbolic representation related linearly (hence "linear encoding"). We show that the state of the system over a finite spatiotemporal domain can be described with exponentially increasing precision by a finite pattern of symbols, and we provide a systematic, lattice Green's function methodology to calculate the frequency (i.e., the measure) of such states.

nlin.CD↗

arXiv2019

Linear chirp instability analysis for ultrafast pulse metrology

Pulse train instabilities have often given rise to confusion in misinterpretation in ultrafast pulse characterization measurements. Most prominently known as the coherent artifact, a partially mode-locked laser with non-periodic waveform may still produce an autocorrelation that has often been misinterpreted as indication for a coherent pulse train. Some modern pulse characterization methods easily miss the presence of a coherent artifact, too. Here we address the particularly difficult situation of a pulse train with chirp-only instability. This instability is shown to be virtually invisible to autocorrelation measurements, but can be detected with FROG, SPIDER, and dispersion scan. Our findings clearly show that great care is necessary to rule out a chirp instability in lasers with unclear mode-locking mechanism and in compression experiments in the single-cycle regime. Among all dynamical pulse train instabilities analyzed so far, this instability appears to be the best hidden incoherence and is most difficult to detect.

physics.optics↗

arXiv2019

100% Reliable Algorithm for Second-Harmonic-Generation Frequency-Resolved Optical Gating

We demonstrate a novel algorithmic approach for the second-harmonic-generation (SHG) frequency-resolved optical gating (FROG) ultrashort-pulse-measurement technique that always converges and, for complex pulses, is also much faster. It takes advantage of the Paley-Wiener Theorem to retrieve the precise pulse spectrum (half the desired information) directly from the measured trace. It also uses a multi-grid approach, permitting the algorithm to operate on smaller arrays for early iterations and on the complete array for only the final few iterations. We tested this approach on more than 25,000 randomly generated complex pulses, yielding SHG FROG traces to which noise was added, and have achieved convergence to the correct pulse in all cases. Moreover, convergence occurs in less than half the time for extremely large traces corresponding to extremely complex pulses with time-bandwidth products up to 100.

physics.optics↗

arXiv2019

100% Reliable Frequency-Resolved Optical Gating Pulse-Retrieval Algorithmic Approach

Frequency-resolved optical gating (FROG) is widely used to measure ultrashort laser pulses, also providing an excellent indication of pulse-shape instabilities by disagreement between measured and retrieved FROG traces. FROG, however, requires -- but currently lacks -- an extremely reliable pulse-retrieval algorithm. So, this work provides one. It uses a simple procedure for directly retrieving the precise pulse spectrum from the measured trace. Additionally, it implements a multi-grid scheme, also quickly yielding a vastly improved guess for the spectral phase before implementing the entire measured trace. As a result, it achieves 100% convergence for the three most common variants of FROG for pulses with time-bandwidth products as high as 100, even with traces contaminated with noise. Here we consider the polarization-gate (PG) and transient-grating (TG) variants of FROG, which measure amplified, UV, and broadly tunable pulses. Convergence occurs for all of the >20,000 simulated noisy PG/TG FROG traces considered and is also faster.

eess.SP↗