We　study　the　usefulness　of　the　broken-line-type　decoherence　in　quantum　Hash　functions　based　on　discrete-time　quantum　walks　on　a　cycle.　We　first　observe　that　the　time-evolution　of　probability　distribution　of　quantum　walks　on　a　cycle　is　of　high　sensitivity　to　such　broken-line-type　decoherence,　increasing　with　the　number　of　steps　in　the　walk.　Based　on　this　observation,　we　further　propose　a　quantum　Hash　function　based　on　the　broken-line　quantum　walk　on　a　cycle.　Numerical　simulation　and　performance　analyses　show　that　decoherence　can　be　useful　to　improve　the　performances　of　quantum　hash　functions　such　as　better　diffusion　and　confusion,　better　collision　resistance　and　more　uniform　distribution　of　hash　values　in　the　hash　space.　Moreover,　our　results　promote　the　practical　use　of　quantum-walk-based　hash　functions　in　realistic　situations.